AUTOMATION HISTORY

Chinese humanoid robots have achieved something that once sounded impossible: beating human athletic records. But the bigger story is not the competition — it is China’s race to build practical robots for the real world. Table of Content Most Read Sponsored TL;DR Chinese humanoid robots have achieved something that sounds like science fiction: they have beaten human athletic benchmarks in sprinting and high jumping. At the World Humanoid Robot Games in Beijing, robots competed across events including running, football, table tennis and industrial challenges designed to test movement and autonomy. One Chinese humanoid robot reportedly completed the 100-metre sprint in 9.39 seconds, faster than Usain Bolt’s 100m world record of 9.58 seconds. Another humanoid robot achieved a standing high jump of 2.88 metres, surpassing the human standing high jump record. The achievements created impressive headlines, but the bigger story is not about robots winning medals. It is about China’s attempt to turn humanoid robots from experimental machines into practical systems that can work in factories, warehouses and everyday environments. This shift is part of the growing physical AI movement, where artificial intelligence moves beyond software and begins controlling machines that interact with the physical world. The Beijing robot games showed how quickly humanoid robotics has progressed. Traditional industrial robots have existed for decades. They can assemble vehicles, move materials and perform repetitive tasks with extreme precision. However, most operate in controlled environments and cannot easily adapt to unexpected situations. Humanoid robots aim to solve a different problem. By using a human-like body design, companies hope these machines can operate in spaces already designed for people, from factories to offices. The latest achievements demonstrate improvements in several important areas: The event was organised around testing these capabilities, with robots competing in multiple categories at the World Humanoid Robot Games official website. But athletic ability is only one part of the challenge. A robot that can run faster than humans is impressive. A robot that can reliably complete useful work every day is far more valuable. The biggest challenge facing humanoid robots is not creating impressive demonstrations. It is reliability. A robot performing one perfect jump on a stage is very different from a robot working inside a factory where conditions constantly change. Businesses need robots that can: This is why the robotics industry is increasingly focused on combining advanced AI models with physical machines. The same AI breakthroughs that created powerful language models are now being applied to robotics. Companies are building systems that allow robots to understand environments, learn tasks and respond more naturally. The rapid growth of AI infrastructure spending shows how companies are investing billions into the computing power required to build more advanced AI systems. That same infrastructure race is now becoming essential for robotics. Humanoid robots need powerful processors and AI models to understand the physical world. China’s progress in humanoid robotics is closely connected to its wider manufacturing strategy. The country already has major advantages in electronics production, batteries, industrial equipment and supply chains. These strengths could help Chinese robotics companies move from prototypes into large-scale manufacturing. The Chinese government has also identified humanoid robots as a strategic technology sector. The Chinese Ministry of Industry and Information Technology (MIIT) has promoted intelligent manufacturing and advanced robotics as part of China’s industrial development plans. The competition is now expanding beyond individual robots. Companies are competing to build complete robotics ecosystems: China is competing with robotics leaders in the US, Japan and Europe, where companies are also investing heavily in humanoid machines. The headlines around robot records can make humanoid robots appear closer to replacing humans than they actually are. Current systems still face major limitations. Battery life remains a challenge. Advanced robots are expensive. Many machines still struggle with simple tasks that humans perform naturally, such as adapting quickly to unfamiliar situations. A human worker can enter a new environment and solve problems with limited instructions. Robots still need extensive training, sensors and carefully designed software. According to research from the International Federation of Robotics (IFR), industrial automation continues to expand globally, but humanoid robots remain an emerging area rather than a fully mature workplace technology. The near-term future is likely to involve humans and robots working together. Factories may use humanoid robots for dangerous, repetitive or physically demanding tasks, while humans focus on supervision, creativity and decision-making. Yes. A Chinese humanoid robot reportedly completed the 100-metre sprint in 9.39 seconds, beating Usain Bolt’s official 9.58-second record. The achievement highlights rapid improvements in robotic movement technology. A humanoid robot achieved a 2.88-metre standing high jump during the Beijing robot games. The result demonstrated progress in robotic balance, motors and motion control. Not yet. Humanoid robots are improving quickly, but companies still need breakthroughs in reliability, battery life, cost and AI decision-making before widespread adoption. Temaz Tra Temaz Tra is an AI and technology news writer focused on the fast-moving tools, platforms, and companies shaping the digital world. He covers artificial intelligence, consumer tech, cybersecurity, software, social media, and the wider impact of emerging technologies on work, business, and everyday life. With a focus on clear reporting and accessible analysis, Temaz helps readers understand complex tech developments without the jargon. His work connects breaking news with practical context, making it easier to follow how AI and digital innovation are changing the way people live, work, and interact online. Read more MemeBurn shares news, reviews, and information about crypto, AI, and technology. Our content is for general information only and should not be taken as financial, investment, legal, or professional advice. Crypto and new technologies can involve risks, and details such as prices, features, and availability may change over time. Some pages may include affiliate links or sponsored content, which may earn MemeBurn a commission, at no additional cost to you. Always do your own research before making any investment or purchase decision. © 2026 MemeBurn. All rights reserved.

news Accueil LNT News Stache Two Human Records Just Got Beaten… By Robots Imagine watching a sports competition where the athletes don’t have muscles, lungs or even a heartbeat. They have batteries. That’s basically what’s happening in China at the World Humanoid Ces dernieres 24h Les galaxies LNT Par LNT LNT Inscrivez-vous à notre newsletter pour recevoir de l’information quotidiennement Recherchez l'actualité qui vous intéresse : Rubriques : Média :

Chinese humanoid robots broke records set by humans, including beating Usain Bolt's 100-meter sprint world record, on the opening day of the Olympics-like World Humanoid Robot Games in Beijing on Saturday. More than 2,000 humanoid robots were participating in the event, the organizer said. The five-day games, now in its second year, are a spectacle demonstrating China's rapid progress in advanced robotics as the technology race with the US heats up, the AP reports, with 51 events and more than 1,000 competitions taking place including running, table tennis, and soccer. The games, which are taking place in the National Speed Skating Oval built for the 2022 Winter Olympics, opened the same week as Beijing held the 2026 World Robot Conference, where companies showcased around 3,000 products, including humanoid robots. China makes the majority of the world's humanoid robots; the US has stepped up scrutiny of robots from the country. At Saturday's games, the organizer and robot makers said Chinese humanoid robots defeated human world records, as hundreds of humanoid robots marched in formation onto the field in a massive display of synchronized coordination. In a 100-meter sprint, a humanoid robot achieved a result of 9.39 seconds, beating the human record of 9.58 seconds set by Jamaican athlete Bolt in 2009. The games, which are taking place in the National Speed Skating Oval built for the 2022 Winter Olympics, opened the same week as Beijing held the 2026 World Robot Conference, where companies showcased around 3,000 products, including humanoid robots. China makes the majority of the world's humanoid robots; the US has stepped up scrutiny of robots from the country. At Saturday's games, the organizer and robot makers said Chinese humanoid robots defeated human world records, as hundreds of humanoid robots marched in formation onto the field in a massive display of synchronized coordination. In a 100-meter sprint, a humanoid robot achieved a result of 9.39 seconds, beating the human record of 9.58 seconds set by Jamaican athlete Bolt in 2009. In a standing high jump, a humanoid robot reached 2.88 meters, well above the 0.95 meters best result by a humanoid in last year's first edition of the games. It surpassed the human high jump record of 2.45 meters set by Cuba's Javier Sotomayor in 1993. Both robots are from Beijing-based X-Humanoid. This year's games—which the organizer said has 16 countries participating, among them Germany, Japan and the US—include other events such as weightlifting and tug of war. Experts say humanoid robots are still mostly used for demonstrations, performances and research, with mass real-world deployment a ways off. Some spectators said they were excited about the humanoid robots' quickly improving abilities. "These sports are perfectly normal for humans, but now robots can do them. I find it amazing," said one. In a standing high jump, a humanoid robot reached 2.88 meters, well above the 0.95 meters best result by a humanoid in last year's first edition of the games. It surpassed the human high jump record of 2.45 meters set by Cuba's Javier Sotomayor in 1993. Both robots are from Beijing-based X-Humanoid. This year's games—which the organizer said has 16 countries participating, among them Germany, Japan and the US—include other events such as weightlifting and tug of war. Experts say humanoid robots are still mostly used for demonstrations, performances and research, with mass real-world deployment a ways off. Some spectators said they were excited about the humanoid robots' quickly improving abilities. "These sports are perfectly normal for humans, but now robots can do them. I find it amazing," said one. Copyright 2026 Newser, LLC. All rights reserved. This material may not be published, broadcast, rewritten, or redistributed. AP contributed to this report.

Written by IPO Zaozhidao Published on 31 Aug 2026 4 mins read Hong Kong-listed UBTech Robotics released its 2026 interim results on August 28, with its full-size humanoid robot business emerging as its largest source of revenue. In the first half, total revenue rose 104.2% year-on-year to RMB 1.27 billion (USD 188.5 million). Revenue from full-size embodied humanoid robot products and services surged 1,445% to RMB 590.3 million (USD 87.6 million) from RMB 38.2 million (USD 5.7 million) a year earlier, accounting for 46.5% of group revenue. Sales volume in the category reached 921 units, up 1,946.7%. The shift in revenue mix also lifted profitability at the gross level. Gross profit increased 160.9% to RMB 566.9 million (USD 84.2 million), while gross margin rose 9.7 percentage points to 44.7%. UBTech attributed the margin improvement mainly to the higher contribution from full-size humanoid robot products and services, which carry higher gross margins. The company nevertheless remained loss-making. Its net loss narrowed 23.0% year-on-year to RMB 338.8 million (USD 50.3 million), while its adjusted EBITDA loss narrowed to RMB 174.1 million (USD 25.8 million) from RMB 321.8 million (USD 47.8 million) a year earlier. UBTech also continued to spend heavily on product development. R&D expenses rose 38.9% to RMB 303.1 million (USD 45 million) in the first half, equivalent to 23.9% of revenue. The company said the increase was mainly driven by continued investment in full-size humanoid robots. Much of that investment is directed toward moving humanoid robots beyond demonstrations and into industrial use. UBTech said its industrial deployments focus on tasks including material handling, loading and unloading, sorting, palletizing, and depalletizing. During the first half, it validated solutions for several of these applications while combining technologies including teleoperation and vision-language-action (VLA) models to address different industrial tasks. The company has also been expanding deliveries of its Walker S2 industrial humanoid robot and deployments of its wheeled Cruzr Y1. Its industrial robots have been tested or deployed in sectors including automotive manufacturing, logistics, and aviation. Airbus, for example, acquired a Walker S2 earlier this year as part of a partnership exploring the use of humanoid robots in aircraft manufacturing. UBTech is broadening its lineup at the same time. In the first half, it launched the Cruzr Y1 for manufacturing and warehousing, the bipedal Walker C1 for commercial services, education, and research, and the U1 series for companionship applications. RELATED ARTICLENewsUBTech’s UWorld U1 tests demand for humanoid robots at homeWritten by Cheng Zi Written by Cheng Zi The three product families reflect UBTech’s attempt to build across industrial, commercial, and eventually household applications rather than rely on a single market. Industrial use remains the most developed of the three. UBTech said it is working to move from individual task validation toward larger deployments in which multiple robots can perform longer sequences of work and collaborate on production lines. Supporting that effort is the company’s artificial intelligence stack. UBTech released Thinker 1.0, its foundation model for embodied intelligence, in the first half. The company said the model ranked first on nine embodied intelligence leaderboards. It also introduced Thinker-WM, a world model designed to help robots model physical environments and anticipate the outcomes of actions. According to UBTech, Thinker-WM ranked first on the Libero embodied intelligence benchmark during the period. Its Thinker-VLA model, meanwhile, combines visual perception, language-based instructions, and action generation. UBTech has been applying VLA technology to industrial handling and loading tasks, where robots need to adjust their actions as objects and surroundings change. The company is also developing BrainNet 2.0 and Co-Agent, technologies intended to coordinate multiple robots. UBTech said the system can divide tasks, schedule operations, share information, and coordinate work among humanoid robots, extending its approach from single-robot autonomy toward multi-robot collaboration. Data is another part of the strategy. UBTech is building humanoid robot data collection and testing centers and a platform designed to connect data collection, annotation, model training, and deployment. It is also developing a simulation platform that can reconstruct scenarios, generate assets and synthetic data, and test models before deployment on physical robots. As of June 30, UBTech held 3,112 granted patents, including 530 overseas patents, up 4.2% from the end of 2025. UBTech is also expanding beyond robot hardware itself. In the first half, the company completed its acquisition of a 43.01% stake in Shenzhen-listed Zhejiang Fenglong Electric, becoming its controlling shareholder. Fenglong’s financial results have been consolidated into UBTech’s accounts since April, contributing RMB 139.2 million (USD 20.7 million) in revenue from garden machinery, automotive, and hydraulic components during the period. Separately, UBTech and Chinese GPU developer MetaX established a joint venture to develop and produce chips for embodied intelligence, extending the company’s involvement further into the hardware stack. UBTech is also working with Siemens on digital tools for robot design and manufacturing. The partnership is intended to support UBTech’s effort to scale production. UBTech has set an annual production target of 10,000 full-size humanoid robots for 2026. For the second half, UBTech plans to continue expanding Walker S industrial models while developing its Walker C and U1 series for commercial, educational, and household applications. It also plans to release an upgraded Thinker-WM 2.0 world model and continue improving its VLA and multi-robot coordination technologies. This article was adapted based on a feature originally written by Stone Jin and published on IPO Zaozhidao. KrASIA is authorized to translate, adapt, and publish its contents. Note: RMB figures are converted to USD at rates of RMB 6.74 = USD 1 based on estimates as of August 31, 2026, unless otherwise stated. USD conversions are presented for ease of reference and may not fully match prevailing exchange rates. Loading... Subscribe to our newsletters KrASIA A digital media company reporting on China's tech and business pulse.

UBTECH Robotics (09880.HK) has reported a short-selling amount of $28.3 million, with a short interest ratio of 18.887%. According to Chief Brand Officer Tan Min, for intelligent manufacturing companies, return on investment is secondary to securing a competitive edge in the emerging productivity landscape. The rise of humanoid robots aids in connecting artificial intelligence to the real world and establishes a new paradigm in intelligent manufacturing. Tan revealed that the company's collaboration with JD.com (09618.HK) involves the launch of ultra-bionic humanoid robots. UBTECH plans to enhance customer experience through flagship stores in major urban centers, focusing on high-spending consumer demographics. The Walker S series, after three iterations, is set to roll out S3 later this year. Looking ahead, UBTECH aims to refine its business model across various sectors, expanding from industrial applications to commercial services and home companionship within the next 3 to 5 years. They expect advancements in artificial intelligence by 2030 will trigger significant growth in industrial, commercial, and household applications. This stock alert was generated using automated technology and GuruFocus financial data to provide readers with timely and accurate market reporting. This content was reviewed by GuruFocus editorial team prior to publication. Please send any questions or comments about this story to [email protected]. We'd love to learn more about your experiences on GuruFocus.com and how we can improve!

NEURA Robotics, a Metzingen-based cognitive robotics startup, today announced that it is acquiring 100% of the shares in Adlatus Robotics, an Ulm-based maker of autonomous cleaning robots. The deal was carried out through NEURA Mobile Robots GmbH, the group’s logistics and automated-vehicle arm, and folds Adlatus’s installed base of cleaning and sweeping robots into NEURA’s broader “Neuraverse” platform. David Reger, founder and CEO of NEURA Robotics, said, “We’re not buying ADLATUS just to add another cleaning robot to our portfolio. We want to give machines like this a new brain. Cleaning is a massive global labour market and, at the same time, a perfect example of what Physical AI can transform. “When a robot doesn’t just navigate its environment but understands what it sees, senses, and needs to do, an entirely new generation of machines emerges. That’s exactly what we’re building the Neuraverse for.” The acquisition comes just weeks after NEURA Mobile Robots agreed to acquire the ACTIVE Shuttle automated guided vehicle business from Bosch Rexroth. That transaction brought Bosch Rexroth’s ACTIVE Fleet Manager and ROKIT navigation software into the NEURA Group, with plans to offer these technologies as tools through the Neuraverse. Founded in 2019, NEURA Robotics builds cognitive robots — including the humanoid 4NE1 and the MAiRA and LARA robotic arms — alongside the software and AI infrastructure that helps these cognitive robots to learn and share skills across the Neuraverse. The company is building a new category of AI infrastructure in which cognitive robots continuously learn, collaborate, and operate across real-world environments through a shared intelligence ecosystem called the Neuraverse. Unlike traditional robotics companies focused on isolated machines or narrow industrial automation, the company states that it combines robotics, AI, sensors, edge compute and large-scale learning infrastructure into one unified platform architecture designed for global deployment. The acquisition spree follows a record capital raise. In June, NEURA raised a Series C financing of up to €1.2 billion ($1.4 billion), which it described as the largest round ever raised by a full-stack robotics company. The funding was backed by Tether, Qualcomm Technologies, Inc., Amazon, NVIDIA, imec.xpand, Bosch, Schaeffler, European Investment Bank, Lingotto Horizon, InterAlpen Partners and others. The company said its existing order book and deployment pipeline already exceeded €864.8 million ($1 billion), and that the fresh capital is earmarked for scaling manufacturing toward several million robots in production by 2030 and expanding its network of “NEURA Gyms”, real-world training environments for cognitive robots. Headquartered in Hamburg, NEURA Mobile Robots develops and delivers mobile transport robotics for production and warehouse logistics at international companies under its ek robotics brand. Headquartered in Ulm, Germany, Adlatus has more than 20 years of experience building autonomous cleaning robots for industry, logistics, healthcare, and commercial and public spaces, with hundreds of systems already deployed in the field. It also builds its own navigation software and has deep knowledge of real-world cleaning operations. Neura says it plans to equip the Adlatus fleet with additional sensor and AI capabilities and gradually connect the machines to the Neuraverse, so that a cleaning robot eventually recognises different surfaces and types of dirt, interprets its surroundings, chooses appropriate cleaning strategies, and learns from deployments rather than simply following a fixed route. The German robotics company says the move doesn’t close the platform to other manufacturers. The company wants the Neuraverse to function as open infrastructure that other makers of cleaning and mobile robots can eventually plug their own machines into. “Physical AI will only truly scale if it becomes an ecosystem,” Reger said. EU-Startups.com is the leading online magazine about startups in Europe. Learn more about us and our advertising options. © Menlo Media S.L. - All rights reserved.

-813.35 -203.60 + 378.00 + 1,186.00 + 4,824.00 -813.35 -203.60 -203.60 + 378.00 + 378.00 + 1,186.00 CEO Tim Cook told employees in an all-hands meeting this month that Apple must win in AI and hinted at the upcoming devices | Photo Credit: GONZALO FUENTES Apple Inc. is plotting its artificial intelligence comeback with an ambitious slate of new devices, including robots, a lifelike version of Siri, a smart speaker with a display and home-security cameras. A tabletop robot that serves as a virtual companion, targeted for 2027, is the centerpiece of the AI strategy, according to people with knowledge of the matter. The smart speaker with a display, meanwhile, is slated to arrive next year, part of a push into entry-level smart-home products. Home security is seen as another big growth opportunity. New cameras will anchor an Apple security system that can automate household functions. The approach should help make Apple’s product ecosystem stickier with consumers, said the people, who asked not to be identified because the initiatives haven’t been announced. Apple shares climbed to a session high on Wednesday after Bloomberg News reported on the plans. The stock was up nearly 2% to $233.70 as of 2:17 p.m. in New York. It’s all part of an effort to restore Apple’s mojo. Its most recent moon-shot project, the Vision Pro headset, remains a sales flop, and the design of its bestselling devices has remained largely unchanged for years. At the same time, Apple has come under fire for missing the generative AI revolution. And OpenAI may even threaten the company’s home turf by developing new AI-driven devices with the help of former Apple design chief Jony Ive. Though Apple is still in the early stages of turning around its AI software, executives see the pipeline of hardware as a key piece of its resurgence — helping it challenge Samsung Electronics Co., Meta Platforms Inc. and others in new categories. A spokesperson for Cupertino, California-based Apple declined to comment. Because the products haven’t been announced, the company’s plans could still change or be scrapped. Many of the initiatives and their timelines rely on Apple’s continued progress in AI-powered software. Chief Executive Officer Tim Cook told employees in an all-hands meeting this month that Apple must win in AI and hinted at the upcoming devices. “The product pipeline — which I can’t talk about — it’s amazing, guys. It’s amazing,” Cook said. “Some of it you’ll see soon. Some of it will come later. But there’s a lot to see.” Beyond the home devices, Apple is preparing thinner and redesigned iPhones for release this year. And further out, it aims to introduce smart glasses, a foldable phone, a 20-year anniversary iPhone and a revamped headset dubbed N100. It’s also planning a large foldable device that melds a MacBook and an iPad. Apple is looking to boost sales after years of slowing growth for its flagship products. It also nixed some expansions into new areas, like self-driving cars, adding pressure to find other sources of revenue. Moreover, the new initiatives will help rebut the idea that the company is no longer innovating like it used to. Bloomberg News first reported last year that Apple was moving forward with a tabletop robotics project, code-named J595, and developing a new smart-home strategy. But now a clearer picture is forming of its push into that market — and what it means for its AI ambitions. The tabletop robot resembles an iPad mounted on a movable limb that can swivel and reposition itself to follow users in a room. Like a human head, it can turn toward a person who is speaking or summoning it, and even seek to draw the attention of someone not facing it. The hope is to bring AI to life in ways that other hardware makers have yet to do. Apple imagines customers placing it on a desk or kitchen counter and using it to get work done, consume media and manage their day. FaceTime calls will also be a key function of the device. During videoconferencing, the display will be able to shift to lock on to people around a room. Apple is testing a feature that turns an iPhone screen into a joystick, letting users move around the robot to show different people or items in a room during video calls. But the hallmark of the device is an entirely new version of the Siri voice assistant that can inject itself into conversations between multiple people. It will be able to engage with users throughout the day and more easily recall information. The idea is for the device to act like a person in a room. It could interrupt a conversation between friends about dinner plans, say, and suggest nearby restaurants or relevant recipes. It’s also being designed to engage in back-and-forth discussions for things like planning a trip or getting tasks done — similar to OpenAI’s voice mode. Apple is planning to put Siri at the center of the device operating system and give it a visual personality to make it feel lifelike. The approach, dubbed Bubbles, is vaguely reminiscent of Clippy, an animated paper clip from the 1990s that served as a virtual assistant in Microsoft Office. Apple has tested making Siri look like an animated version of the Finder logo, the iconic smiley face representing the Mac’s file management system. A final decision on its appearance hasn’t been made, with designers considering ideas that veer closer to Memoji, the playful characters that represent Apple user accounts. Device prototypes use a roughly 7-inch horizontal display, approaching the size of an iPad mini. The motorized arm can extend the display away from the base roughly half a foot in any direction. Some people familiar with the product call it the “Pixar Lamp,” referring to the animated film company’s famous logo. Apple has previously disclosed some research in this area: It published a paper in January detailing a light fixture that uses robotics to move around. Apple has multiple teams across its AI, hardware, software and interface design groups tackling the project. The work is being led in part by Kevin Lynch, who previously oversaw a push into smart watch software and cars. The technology giant is developing several other robots. It has teams exploring a mobile bot with wheels — something akin to Amazon.com Inc.’s Astro — and has loosely discussed humanoid models. Apple has a group actively developing a large mechanical arm for use in manufacturing facilities or handling tasks in the back of retail stores, a move that could potentially replace some staff. Such a robot, code-named T1333, remains several years away. Charismatic The smart-home push includes a standalone display poised to launch by the middle of next year. That device, code-named J490, is a stripped-down variant of the robot, lacking the arm and conversational Siri — at least to start. It will still have home control, music playback, note taking, web browsing and videoconferencing. It may also include the new Siri visual interface. Both the smart display and tabletop robot will run a new operating system dubbed Charismatic, which is designed to be used by multiple people. The interface largely centers on clock faces and widgets — small software features that are typically dedicated to specific tasks. Charismatic, which was previously known as Pebble and Rock earlier in development, blends the approach of the Apple TV and Apple Watch operating systems. It offers features like multiuser modes and clock-face themes, such as one based on Snoopy, the beagle from the Peanuts comic strip. The devices are meant to be easily shared: They include a front-facing camera that can scan users’ faces as they walk toward it and then automatically change the layout, features and content to the preferences of that person. Some versions of the software use circular app icons and feature a hexagonal grid of apps. Apple is planning to include many of its core apps, including the calendar, camera, music, reminders and notes software. But the interface will be heavily reliant on voice interaction and widgets, rather than jumping in and out of apps. Though the device will have a touch screen, the primary input method will be Siri and an upcoming upgrade to a feature dubbed App Intents. That software lets users precisely control the interface and applications via voice. The hardware itself looks similar to a Google Nest Hub but is shaped like a square, with thin black or white bezels and rounded corners. The non-robotic 7-inch display sits on a half-dome-shaped base, which includes some of the electronics and is perforated around the bottom edges for speakers and microphones. It can also be mounted on a wall. The launch will mark the first time Apple is making a serious push into the smart home and comes nearly a decade after Amazon and Alphabet Inc.’s Google started shipping smart speakers with screens. The home is a critical space for Apple to target, especially as more users consume content from the living room and automate household functions. Apple has long had a strong foothold in mobile devices and quickly became a player in the automotive industry via CarPlay — but that success hasn’t followed into the smart home. Though the company launched HomeKit for controlling third-party devices in 2014, it has had limited success with its own HomePod speakers. Linwood and Glenwood Core to the new home devices — and current products like iPhones and iPads — is an overhaul to the underpinnings of Siri. Engineers are working on a version code-named Linwood with an entirely new brain built around large language models — the foundation of generative AI. The goal is to tap into personal data to fulfill queries, an ability that was delayed due to hiccups with the current version. That new software, known internally as LLM Siri, is planned for release as early as next spring, Bloomberg News has reported. But work is going even further: Apple is preparing a visually redesigned assistant for iPhones and iPads that will also debut as early as next year. Craig Federighi, senior vice president of software engineering, hinted at a bigger-than-anticipated overhaul in an internal meeting with employees this month. “The work we’ve done on this end-to-end revamp of Siri has given us the results we needed,” he said, adding that “this has put us in a position to not just deliver what we announced, but to deliver a much bigger upgrade than we envisioned.” He said that “there is no project people are taking more seriously.” Linwood is based on technology developed by the Apple Foundation Models team, but the company has a competing project dubbed Glenwood as well that would power Siri with outside technology. A final decision hasn’t been made on which models will be used, but Apple has been testing Anthropic PBC’s Claude for this purpose. Mike Rockwell, the former Vision Pro chief who was put in charge of Siri earlier this year, is overseeing both the Linwood and Glenwood efforts. During development of the tabletop robot, Apple engineers have made heavy use of ChatGPT and Google Gemini to build and test features. Within Apple’s AI and Siri teams as a whole, software developers are increasingly using third-party systems as part of their development process. Ring Competitor Apple is working on a camera, code-named J450, designed for home security, detecting people and automating tasks. The device will be battery-powered and could last from several months to a year on a single charge, on par with rival products. The device has facial recognition and infrared sensors to determine who is in a room. Apple believes users will place cameras throughout their home to help with automation. That could mean turning lights off when someone leaves a room or automatically playing music liked by a particular family member. Apple is planning to develop multiple types of cameras and home-security products as part of an entirely new hardware and software lineup. The goal is to compete with Amazon Ring, Google Nest and Roku Inc., capitalizing on its customer loyalty to sell more products. It has also tested a doorbell that uses facial recognition technology to unlock a door. Apple already sells iCloud+ subscriptions with online storage for security footage, but they’re aimed at third-party cameras. ©2025 Bloomberg L.P. Published on August 14, 2025 Copyright© 2026, THG PUBLISHING PVT LTD. or its affiliated companies. All rights reserved. BACK TO TOP Comments have to be in English, and in full sentences. They cannot be abusive or personal. Please abide by our community guidelines for posting your comments. We have migrated to a new commenting platform. If you are already a registered user of TheHindu Businessline and logged in, you may continue to engage with our articles. If you do not have an account please register and login to post comments. Users can access their older comments by logging into their accounts on Vuukle. Terms & conditions | Institutional Subscriber

Der letzte Führerscheinneuling… …ist bereits geboren. Wie Waymo, Tesla, Zoox & Co unsere automobile Gesellschaft verändern, und Mobilität sicherer, günstiger und in städtischen wie auch in ländlichen Regionen zugänglicher machen werden. Stellenausschreibungen von Tesla für den Standort Holzgerlingen südwestlich von Stuttgart suchen nach Ingenieuren, die an den Produktionsmaschinen für Getriebe für den Tesla Optimus arbeiten sollen. Bislang sind zwei Produktionsstandorte für den Teslas humanoiden Roboter Optimus bekannt. Einer ist in Fremont, der anstelle der bisherigen Model S und Model X-Produktion die Werkhalle mit einer Jahreskapazität von einer Million Humanoider einnehmen soll, der andere ist Austin, wo gerade eine neue Fabrik mit einer Jahreskapazität von 10 Millionen errichtet wird. Die Ausschreibungen für die Roboteringenieure in Holzgerlingen scheinen bei Tesla Autmoation zu gelten. Diese Einheit errichtet für Tesla weltweit die Produktionslinien. Bedeutet die Suche nach Verstärkung ausgerechnet in Deutschland, dass Tesla (auch) einen Produktionsstandort für den Tesla Optimus in Deutschland plant? Dieser Beitrag ist auch auf Englisch erschienen. Zeige alle Beiträge von Mario Herger Δ

Robots are protesting for human rights against AI. Yes, it might sound unusual, but around 30 robots in Poland have taken to the streets to raise concerns about AI regulation and the growing threat of automation to human jobs. The unusual protest took place outside Poland’s Digital Affairs Ministry in Warsaw earlier this week, with humanoid robots and robot dogs joining the demonstration and calling for stronger rules around AI. The protest was staged by an initiative called Democratism, with robots marching in circles, waving flags and broadcasting slogans through loudspeakers. According to the organisers, the demonstration was intended to start a wider debate about how AI and robotisation are affecting the labour market, while also showing that the technology is no longer a distant prospect. Among those at the demonstration was Agibot A3, an AI-powered humanoid robot and different types of machines, from bipedal humanoid robots to dog-like quadrupeds. Videos from the demonstration, which have been circulating online, show the robots holding flags and signs as loudspeakers played slogans such as “Defend workplaces”, “Time for rules” and “Don’t wait, regulate”. Human organiser Grzegorz Kulis said the protest was intended as a warning about the speed at which AI and robotics are developing. “We particularly want to address the issue of the potential replacement of people both by humanoid robots and by artificial intelligence in cognitive work,” he told AFP. “If we don’t react now, we may soon have a problem,” he added, calling for “rules and regulations that would serve as a shield protecting workers”. Kulis also explained why robots were made the centrepiece of the demonstration. According to him, their presence was meant not only to attract the attention of passers-by, but also to show what the machines can already do and underline the urgency of the issue. The protest comes at a time when Poland is putting new rules around AI into place as part of the European Union’s wider push to regulate the technology. The EU AI Act, which entered into force in 2024, follows a risk-based approach, with stricter requirements for AI systems considered more likely to cause harm. Most of its rules began applying from August 2026, with enforcement now underway for areas including prohibited AI practices, transparency requirements and general-purpose AI models. Poland has also passed its own legislation to put the EU rules into practice. In July, President Karol Nawrocki signed the AI systems law, creating a national framework for supervising AI and a new Commission for the Development and Safety of Artificial Intelligence. The body will oversee the application of AI rules in Poland and handle complaints about potentially harmful AI systems.- EndsPublished By: Divya BhatiPublished On: Sep 8, 2026 12:44 ISTAlso Read | LG smart TVs may scan your network and record audio even when screen is off, report saysAlso Read | Robot making robots? China's Xpeng launches automated production line to build advanced humanoidsAlso Read | AI helps create drug that may slow down ageing by 6 years in few weeks Ruckus In Tamil Nadu Assembly, DMK MLAs Evicted, Protest Over Chief Minister's Remarks | Tamil Nadu CM Issues Clarification on Assembly Gesture, Denies Taunt at MK Stalin | UAE president warned Netanyahu of Hamas operation before October 7: Haaretz | PM Event Security Scare: Imposter Nabbed In BKC, Claims Con Job, Remanded In Custody | Bangladesh President seeks early talks with India on pending bilateral issues | Leopard hit by car attacks forest official during rescue in Karnataka | Bengaluru entrepreneur spends 4 months in ICU, shares what it taught him | Abdullahs stand for full Vande Mataram weeks after Congress objects | A letter, a mystery and 3 generations: Inside Atima Mankotia’s latest book | China's DeepSeek goes on hiring spree, 150 vacancies currently open for engineers | Singapore raises ministers' pay after 15 years to attract political talent | PM Modi says degrees are no longer enough, skills matter more for India's youth | Unacceptable: Kremlin hits US sanctions, says Russia doesn't seek de-dollarisation | Can’t keep it lingering: SC pulls up Odisha over delay in Dara Singh remission plea | Mumbai Indians star goes berserk in Punjab T20 League, smashes 173 off just 64 balls

Vanguard News September 7, 2026 Humanoid and quadruped robots are seen during a demonstration calling for the regulation of artificial intelligence development in Warsaw on September 7, 2026. Waving flags, “chanting” slogans, and marching in circles — around thirty robots took to the streets to raise awareness for AI regulation in front of the digital affairs ministry on September 7, 2026. (Photo by Sergei GAPON / AFP) Waving flags, “chanting” slogans, and marching in circles: around 30 robots took to the streets of Warsaw on Monday to campaign for AI regulation in front of the Polish digital affairs ministry. The protest was organised by Democratism, an initiative that, according to its website, aims to launch debate “on how AI and robotisation affect the labour market”. “The robots present here are meant to show that this technology is already a reality,” Agibot A3, one of the AI-powered humanoid robots, told AFP. Chants of: “Defend workplaces”, “Time for rules”, and “Don’t wait, regulate” played on loudspeakers as the robots — which ranged from bipedal humanoids to dog-like quadripeds — demonstrated. “We particularly want to address the issue of the potential replacement of people both by humanoid robots and by artificial intelligence in cognitive work,” organiser Grzegorz Kulis told AFP. “If we don’t react now, we may soon have a problem,” he added, emphasising the need for “rules and regulations that would serve as a shield protecting workers”. Aside from attracting the attention of several passers-by, Kulis said the use of robots was meant to show their existing capabilities and thereby, the urgency of the situation. “These robots can already move around freely, they’re already ‘protesting'”, he said. The marching robots were met with a visit from Poland’s Digital Affairs Minister Krzysztof Gawkowski, who spoke to Kulis on site. “Uncontrolled (AI) models, implemented into humanoids that will also keep evolving as robotisation moves forward, are a major problem,” he said. Kulis, who is now the CIO of a robotics company and CEO of an employment agency, served two terms on the Labor Market Council at the Ministry of Family, Labor and Social Policy. “We feel responsible for starting to inform society about the rapid and dynamic technological changes which personally frighten us,” he told AFP. In July, Poland’s conservative-nationalist president Karol Nawrocki signed into law a bill put forth by the ministry putting the EU’s AI Act into practice. It created a national authority capable of enforcing existing EU regulations, which had already formally come into force. WhatsApp Channel Follow Vanguard Telegram Channel Follow Vanguard Comments expressed here do not reflect the opinions of Vanguard newspapers or any employee thereof. Subscribe to our digital E-Editions here, and enjoy access to the exact replica of Vanguard Newspapers publications.

Humans still performed 75 to 80% of tasks in even the most advanced factories, Zeeshan Zia, CEO and Co-Founder of Retrocausal, told a Sustainability Day 2024 panel with Eve Psalti, Senior Director at Microsoft, and Michael Kuehne-Schlinkert, Co-Founder and CEO of Katulu. The panel covered where AI creates value beyond robotics and how to scale it across sites. Robots handle welding, painting, and pick-and-place, but the remaining 75 to 80% of factory tasks still depend on people. The U.S. Bureau of Labor Statistics identifies industrial engineering as one of the toughest fields to recruit for. AI fills this gap by making workers more productive through just-in-time training, augmented reality guidance, and process optimization that recommends assembly layout adjustments or automates robot programming. “AI should address human work, providing tools that improve cycle times, reduce errors, and adapt to labor shortages,” Zia said. AI models also anticipate failures in predictive maintenance and support as-a-service business models. Computer vision catches defects earlier in quality control, saving resources and improving efficiency. The AI ecosystem offers three categories of models. Traditional AI services include computer vision, text analytics, and speech recognition for tasks like defect detection and customer service. Generative AI models handle complex, multi-step tasks simultaneously, such as transcribing calls, analyzing sentiment, and summarizing conversations. Multimodal AI models process text, images, video, and audio inputs for richer and more versatile applications. The right choice depends on the problem: smaller, specialized models drive faster return on investment for targeted use cases, while generative AI fits complex, multi-step tasks. Successful AI deployment starts with identifying specific pain points and focusing on high-priority inefficiencies. Clean, labeled, and relevant data is critical, since training AI models depends on data quality. A center of excellence helps share best practices across departments and accelerate adoption, and employees need training to use AI tools effectively and foster a culture of improvement. Manufacturers should also address bias, transparency, and traceability to ensure ethical implementation. Proof-of-concept projects often fail because of misalignment between ground-level operations and leadership goals. Champions of new technology need to understand internal buying processes, align expectations, and define clear ROI metrics before deployment begins. Many POCs are also designed for specific configurations and do not generalize well across different processes or factories. Help Build the Industry Benchmark: Take the 2027 Industrial Data & AI Readiness Survey and Get Early Access to the Report AI augments factory workers through tools that improve cycle times, reduce errors, and adapt to labor shortages. According to Retrocausal, applications include just-in-time training, augmented reality guidance, and process optimization that recommends assembly layout adjustments or automates robot programming. These tools target the 75 to 80% of factory tasks that humans still perform in even the most advanced factories. Manufacturing AI models fall into three categories according to Microsoft. Traditional AI services include computer vision, text analytics, and speech recognition for tasks like defect detection and customer service. Generative AI models handle complex, multi-step tasks such as transcription, sentiment analysis, and summarization simultaneously. Multimodal models process text, images, video, and audio inputs for richer and more versatile applications. Manufacturers should start by identifying specific pain points and focusing on high-priority inefficiencies, according to a Microsoft senior director at the panel. Data readiness is the second priority: training data must be clean, labeled, and relevant. Establishing a center of excellence helps share best practices across departments and accelerate adoption. Employee training and responsible AI practices, including bias, transparency, and traceability, complete the preparation. AI proof-of-concept projects in manufacturing commonly fail because of misalignment between ground-level operations and leadership goals, according to Retrocausal and Katulu panelists. Champions of new technology need to understand internal buying processes, align expectations, and define clear ROI metrics before deployment. Many POCs are also designed for specific configurations that do not generalize well across different processes or factories. Related from IIoT World Sources: 1. Sustainability Day 2024 panel: “AI-Driven Process Optimization: Achieving Faster Turnarounds and Higher Margins,” sponsored by Retrocausal This article is based on a panel discussion with Zeeshan Zia, CEO and Co-Founder of Retrocausal, Eve Psalti, Senior Director at Microsoft, and Michael Kuehne-Schlinkert, Co-Founder and CEO of Katulu, moderated by Hamish Mackenzie at Sustainability Day 2024. AI tools were used to help summarize and organize the content. Reviewed and edited by the IIoT World editorial team. P&G evaluates 20 AI models every day to forecast six weeks of demand for Charmin and Bounty, Jaimie McIntyre Horstman,… Cognex reported that Schneider Electric doubled its production yield and eliminated most false rejects after deploying the OneVision AI vision… Between 80% and 95% of organizations report limited financial value from their AI investments, and only 11% of CFOs have… © 2017-2026 IIoT World. All articles submitted by our contributors do not constitute the views, endorsements or opinions of IIoT-World.com. Sep 09–10 / Virtual / Free LIVE PANELS on agentic AI, edge deployment, digital twins, and measurable ROI on the plant floor. We are running our annual survey. Your experience becomes the benchmark the industry cites next year, and it is yours, free.

We’re so glad you’re here. You can expect all the best TNS content to arrive Monday through Friday to keep you on top of the news and at the top of your game. Check your inbox for a confirmation email where you can adjust your preferences and even join additional groups. Follow TNS on your favorite social media networks. Become a TNS follower on LinkedIn. Check out the latest featured and trending stories while you wait for your first TNS newsletter. You could have a mechanical duck waddling through your home before Christmas. Hugging Face‘s Pollen Robotics on Thursday opened pre-orders for the Microduck, a $399 (introductory) bipedal duck-adjacent robot that can walk, waddle, and use roller-skates(!), with a beak to pick up objects. And when it falls, it can get back up, too. The company expects to make the first deliveries of the Microduck before Christmas. It’s available in North America and Europe. Microduck is the follow-up to Reachy Mini, the desktop robot Hugging Face and Pollen launched last year, which was stationary and focused on interacting with humans. Pollen also still sells Reachy 2, a far larger and pricier humanoid aimed at research labs. Microduck, the team writes in its announcement, is meant to focus on action. “How do you teach a robot to move? How do you train a behavior in simulation, transfer it to real hardware, see what went wrong, and try again? What changes when the robot can leave the desk, carry something, fall over, and recover? It is an ideal platform for developers who want to train physical behaviors, experiment with reinforcement learning, and test how AI moves from simulation into the real world,” the team writes. And indeed, Microduck is not just a 25cm-tall toy. It’s an open-source platform with an SDK, virtual training environment, and reinforcement learning scripts to help developers train the robot to perform new tasks. The code is Apache 2.0, though the hardware design files are licensed non-commercially, so nobody is building and selling a clone. There’s also a full simulator for those of us who just want to play with a duck robot, but if you do buy one, you’ll also get a game controller to control the robot on the fly, too. But if you want to go deep, you can use the physics simulator to teach the robot new movements. On the project’s GitHub page, the team notes how to train the duck’s walking policy across 4,096 virtual ducks in parallel, for example, which results in a usable gait in one to two hours. The repo registers 13 task families in all, including a forward roll and six built around a set of passive wheels that go under the feet. To train the robot, you’ll need an Nvidia GPU, or you can train it on Hugging Face’s own infrastructure. That’s not incidental. The simulator the ducks run in is MuJoCo Warp, built on Nvidia’s Warp framework, and mjlab, the training framework underneath, reimplements the API of Nvidia’s own Isaac Lab. Out of the box, the robot comes with seven trained moves, including walking, sitting and standing, kicking, grabbing objects with its beak, roller skating, and getting back up off the ground. As for the hardware, the robot will weigh in at about 800 grams and will be powered by a Rockchip RK3566 with AI accelerator. That’s basically a quad-core Arm Cortex-A55 with a Mali GPU. But now that Nvidia is reportedly in talks to acquire Hugging Face, in a deal first reported by The Information, I would expect a future version to use a slightly more powerful Nvidia-made chip. It features 1GB of on-board memory and 32 GB of storage. What’s more important, though, is its set of sensors. There’s a single front camera, a small LiDAR sensor with an 8×8 time-of-flight matrix, and 2 inertial measurement units. The camera and LiDAR let it see and place objects around it, while the IMUs keep track of its orientation and balance. The team is still working out what the final camera resolution and LiDAR range will be. Pollen Robotics will sell a few accessories as well, including, for example, a Charger Pack for $39 with two batteries and a charger, and a Dev Pack for $119 with three spare motors, five motor cables, two batteries, a dual charger, ten NFC tags, Hugging Face credit, screws, and a screwdriver. The robot also comes with microphones and a speaker. One interesting note here: when you first turn the robot on, it generates its own signature sound that is different from any other Microduck. The robot doesn’t speak, though, as the team notes, the Microduck “communicates through weird little sounds, closer to a creature than an assistant.” The team says having several of them together is what really makes the robots come alive. “Races, football, or simply robots reacting to one another immediately make the experience feel more alive. For developers, it also creates a practical way to explore multi-robot behaviors without a room full of expensive hardware,” they write. At the end of the day, this is also just a fun project, and in this depressing world, we all deserve some ducking fun every now and then. Community created roadmaps, articles, resources and journeys for developers to help you choose your path and grow in your career.

Tesla Cybercab is coming to Asia this month as US service officially begins Tesla Cybertruck targets job site crews with new Tailgate Utility Track and Bed Gear Box accessory Tesla opens Cybercab rides to the public, with no steering wheel or pedals Tesla hints its already prepping for Cybercab fleet orders Tesla Robotaxi riders will face the best dilemma when booking a ride Tesla Cybercab is coming to Asia this month as US service officially begins Tesla opens Cybercab rides to the public, with no steering wheel or pedals Tesla hints its already prepping for Cybercab fleet orders Tesla Robotaxi riders will face the best dilemma when booking a ride Tesla Cybercab sightings broaden well outside of Austin with autonomy in focus SpaceX tells the FCC that Starship Flight 14 is going to orbit SpaceX would not exist if this crucial early launch failed, Musk says OpenAI cites distrust of SpaceX in decision to drop Cursor partnership SpaceX announces new Starbase for ‘thousands of Starship launches annually’ Elon Musk just set a condition that could end Falcon 9 as we know it SpaceX would not exist if this crucial early launch failed, Musk says Elon Musk’s Grok can basically control anything in your Tesla now Tesla is fixing Full Self-Driving’s pothole problem Elon Musk says he knows how to save Earth for a billion years OpenAI cites distrust of SpaceX in decision to drop Cursor partnership Published on By Tesla is starting to prepare for the launch of the Optimus robot in its smartphone app, new coding strings show. Elon Musk has referred to Optimus as what will be the greatest-selling product of any kind of all time, and now, Tesla is getting ready for its launch. Tesla’s smartphone app had several first-time mentions of the Optimus program, according to Tesla App Updates, who intially reported on the appearance. Here’s what they found: A Dedicated “Robot” Phone Key Authentication Tesla is working on a Bluetooth Low Energy, or BLE, authentication that is specifically for robots. This does not only apply to Optimus, though, as Robotaxi, which is Tesla’s autonomous ride-hailing platform, might also identify vehicles within the fleet as robots as well. Tesla shows rapid teardown of Model S and X lines, paving the way for Optimus at Fremont Essentially, pairing your phone as a key to anything Tesla identifies as a robot to a “whitelist” of authorized devices. Optimus, Robotaxi, or other products that fall into this category will only respond if the device trying to communicate with it is authorized. This is a great security feature that will eliminate at least face-value and low-level threats.Advertisement - Home Data Collection and System Alerts This appears to be somewhat of a neural network for Optimus within your house. There will be a dedicated screen that asks for consent to collect both video and spatial data while Optimus performs in-home tasks. Everything from vacuuming, washing dishes, dusting, and other activities will be tracked. There will also be a comprehensive alert system that will track everything from low battery to mechanical issues. Other Changes Most of the changes tracked in this particular app update are related to Tesla’s 2026 Summer Update, and include things such as image assets for new features, a preview of the new custom wraps feature, and other unique features. You can check out our coverage on what is included with the 2026 Summer Update here: Tesla reveals 2026 Summer Update with crazy fixes to Nav and more Advertisement - Tesla’s smartphone app had several first-time mentions of the Optimus program, according to Tesla App Updates, who intially reported on the appearance. Here’s what they found: A Dedicated “Robot” Phone Key Authentication Tesla is working on a Bluetooth Low Energy, or BLE, authentication that is specifically for robots. This does not only apply to Optimus, though, as Robotaxi, which is Tesla’s autonomous ride-hailing platform, might also identify vehicles within the fleet as robots as well. Tesla shows rapid teardown of Model S and X lines, paving the way for Optimus at Fremont Essentially, pairing your phone as a key to anything Tesla identifies as a robot to a “whitelist” of authorized devices. Optimus, Robotaxi, or other products that fall into this category will only respond if the device trying to communicate with it is authorized. This is a great security feature that will eliminate at least face-value and low-level threats.Advertisement - Home Data Collection and System Alerts This appears to be somewhat of a neural network for Optimus within your house. There will be a dedicated screen that asks for consent to collect both video and spatial data while Optimus performs in-home tasks. Everything from vacuuming, washing dishes, dusting, and other activities will be tracked. There will also be a comprehensive alert system that will track everything from low battery to mechanical issues. Other Changes Most of the changes tracked in this particular app update are related to Tesla’s 2026 Summer Update, and include things such as image assets for new features, a preview of the new custom wraps feature, and other unique features. You can check out our coverage on what is included with the 2026 Summer Update here: Tesla reveals 2026 Summer Update with crazy fixes to Nav and more Advertisement - Tesla is working on a Bluetooth Low Energy, or BLE, authentication that is specifically for robots. This does not only apply to Optimus, though, as Robotaxi, which is Tesla’s autonomous ride-hailing platform, might also identify vehicles within the fleet as robots as well. Tesla shows rapid teardown of Model S and X lines, paving the way for Optimus at Fremont Essentially, pairing your phone as a key to anything Tesla identifies as a robot to a “whitelist” of authorized devices. Optimus, Robotaxi, or other products that fall into this category will only respond if the device trying to communicate with it is authorized. This is a great security feature that will eliminate at least face-value and low-level threats.Advertisement - Home Data Collection and System Alerts This appears to be somewhat of a neural network for Optimus within your house. There will be a dedicated screen that asks for consent to collect both video and spatial data while Optimus performs in-home tasks. Everything from vacuuming, washing dishes, dusting, and other activities will be tracked. There will also be a comprehensive alert system that will track everything from low battery to mechanical issues. Other Changes Most of the changes tracked in this particular app update are related to Tesla’s 2026 Summer Update, and include things such as image assets for new features, a preview of the new custom wraps feature, and other unique features. You can check out our coverage on what is included with the 2026 Summer Update here: Tesla reveals 2026 Summer Update with crazy fixes to Nav and more Advertisement - Tesla shows rapid teardown of Model S and X lines, paving the way for Optimus at Fremont Essentially, pairing your phone as a key to anything Tesla identifies as a robot to a “whitelist” of authorized devices. Optimus, Robotaxi, or other products that fall into this category will only respond if the device trying to communicate with it is authorized. This is a great security feature that will eliminate at least face-value and low-level threats.Advertisement - Home Data Collection and System Alerts This appears to be somewhat of a neural network for Optimus within your house. There will be a dedicated screen that asks for consent to collect both video and spatial data while Optimus performs in-home tasks. Everything from vacuuming, washing dishes, dusting, and other activities will be tracked. There will also be a comprehensive alert system that will track everything from low battery to mechanical issues. Other Changes Most of the changes tracked in this particular app update are related to Tesla’s 2026 Summer Update, and include things such as image assets for new features, a preview of the new custom wraps feature, and other unique features. You can check out our coverage on what is included with the 2026 Summer Update here: Tesla reveals 2026 Summer Update with crazy fixes to Nav and more Advertisement - Essentially, pairing your phone as a key to anything Tesla identifies as a robot to a “whitelist” of authorized devices. Optimus, Robotaxi, or other products that fall into this category will only respond if the device trying to communicate with it is authorized. This is a great security feature that will eliminate at least face-value and low-level threats.Advertisement - Home Data Collection and System Alerts This appears to be somewhat of a neural network for Optimus within your house. There will be a dedicated screen that asks for consent to collect both video and spatial data while Optimus performs in-home tasks. Everything from vacuuming, washing dishes, dusting, and other activities will be tracked. There will also be a comprehensive alert system that will track everything from low battery to mechanical issues. Other Changes Most of the changes tracked in this particular app update are related to Tesla’s 2026 Summer Update, and include things such as image assets for new features, a preview of the new custom wraps feature, and other unique features. You can check out our coverage on what is included with the 2026 Summer Update here: Tesla reveals 2026 Summer Update with crazy fixes to Nav and more Advertisement - This is a great security feature that will eliminate at least face-value and low-level threats.Advertisement - Home Data Collection and System Alerts This appears to be somewhat of a neural network for Optimus within your house. There will be a dedicated screen that asks for consent to collect both video and spatial data while Optimus performs in-home tasks. Everything from vacuuming, washing dishes, dusting, and other activities will be tracked. There will also be a comprehensive alert system that will track everything from low battery to mechanical issues. Other Changes Most of the changes tracked in this particular app update are related to Tesla’s 2026 Summer Update, and include things such as image assets for new features, a preview of the new custom wraps feature, and other unique features. You can check out our coverage on what is included with the 2026 Summer Update here: Tesla reveals 2026 Summer Update with crazy fixes to Nav and more Advertisement - This appears to be somewhat of a neural network for Optimus within your house. There will be a dedicated screen that asks for consent to collect both video and spatial data while Optimus performs in-home tasks. Everything from vacuuming, washing dishes, dusting, and other activities will be tracked. There will also be a comprehensive alert system that will track everything from low battery to mechanical issues. Other Changes Most of the changes tracked in this particular app update are related to Tesla’s 2026 Summer Update, and include things such as image assets for new features, a preview of the new custom wraps feature, and other unique features. You can check out our coverage on what is included with the 2026 Summer Update here: Tesla reveals 2026 Summer Update with crazy fixes to Nav and more Advertisement - There will also be a comprehensive alert system that will track everything from low battery to mechanical issues. Other Changes Most of the changes tracked in this particular app update are related to Tesla’s 2026 Summer Update, and include things such as image assets for new features, a preview of the new custom wraps feature, and other unique features. You can check out our coverage on what is included with the 2026 Summer Update here: Tesla reveals 2026 Summer Update with crazy fixes to Nav and more Advertisement - Most of the changes tracked in this particular app update are related to Tesla’s 2026 Summer Update, and include things such as image assets for new features, a preview of the new custom wraps feature, and other unique features. You can check out our coverage on what is included with the 2026 Summer Update here: Tesla reveals 2026 Summer Update with crazy fixes to Nav and more Advertisement - You can check out our coverage on what is included with the 2026 Summer Update here: Tesla reveals 2026 Summer Update with crazy fixes to Nav and more Advertisement - Tesla reveals 2026 Summer Update with crazy fixes to Nav and more Advertisement - Joey has been a journalist covering electric mobility at TESLARATI since August 2019. In his spare time, Joey is playing golf, watching MMA, or cheering on any of his favorite sports teams, including the Baltimore Ravens and Orioles, Miami Heat, Washington Capitals, and Penn State Nittany Lions. You can get in touch with joey at joey@teslarati.com. He is also on X @KlenderJoey. If you're looking for great Tesla accessories, check out shop.teslarati.com Tesla Asia says Cybercab will be on display in Hong Kong, Tokyo, Beijing and Shanghai this month. Published on By Tesla’s Cybercab is heading to Asia. The official Tesla Asia account posted on X Thursday, inviting Cybercab fans to “Come experience the future of autonomy in Hong Kong, Tokyo, Beijing & Shanghai.” The post went up within hours of Tesla’s own Cybercab milestone in Texas, where the company said Thursday it had begun offering rides in across Austin. Exact dates and venues for the Asia tour haven’t been released yet, though Tesla Hong Kong replied to the announcement with “Cybercab will be on display in Hong Kong soon,” while Tesla Japan’s response pointed fans to a sign up page for updates. Neither post mentions test rides or a service area, and nothing so far suggests Tesla is launching Robotaxi operations in any of the four cities. Based on how Tesla has run past Cybercab tours, in Europe in late 2024 and at US shopping centers that same December, the Asia stops are almost certainly static displays at Tesla stores or public venues as a means to stimulate buzz for its future driverless ride-hailing service in the big cities. Cybercab will be on display in Asia this month! Come experience the future of autonomy in Hong Kong, Tokyo, Beijing & Shanghai. pic.twitter.com/wGmmEastfX — Tesla Asia (@Tesla_Asia) September 4, 2026 The timing lines up with Tesla’s only prior Cybercab appearance in the region, a booth at the China International Import Expo in Shanghai last November, which Teslarati covered at the time. At that event, Tesla’s regional general manager for Shanghai framed the car as evidence of the company’s broader mission, a message Tesla has since formalized in its Master Plan Part IV, which states that “autonomous vehicles have the capacity to dramatically improve the affordability, availability and safety of transportation while reducing pollution, particularly in our increasingly dense global cities.” The same document is where Tesla lays out its “sustainable abundance” framing for Cybercab and Optimus alike, describing the two as the hardware behind an AI driven push to cut the cost of transportation and labor at scale. Whether Cybercab actually operates as a robotaxi anywhere in Asia remains an open question, considering China has already pushed an autonomous ride-hailing market that’s run on homegrown players like Baidu’s Apollo Go and Pony AI. For now, the four city tour reads as a marketing push timed to Austin’s momentum.Advertisement - Exact dates and venues for the Asia tour haven’t been released yet, though Tesla Hong Kong replied to the announcement with “Cybercab will be on display in Hong Kong soon,” while Tesla Japan’s response pointed fans to a sign up page for updates. Neither post mentions test rides or a service area, and nothing so far suggests Tesla is launching Robotaxi operations in any of the four cities. Based on how Tesla has run past Cybercab tours, in Europe in late 2024 and at US shopping centers that same December, the Asia stops are almost certainly static displays at Tesla stores or public venues as a means to stimulate buzz for its future driverless ride-hailing service in the big cities. Cybercab will be on display in Asia this month! Come experience the future of autonomy in Hong Kong, Tokyo, Beijing & Shanghai. pic.twitter.com/wGmmEastfX — Tesla Asia (@Tesla_Asia) September 4, 2026 The timing lines up with Tesla’s only prior Cybercab appearance in the region, a booth at the China International Import Expo in Shanghai last November, which Teslarati covered at the time. At that event, Tesla’s regional general manager for Shanghai framed the car as evidence of the company’s broader mission, a message Tesla has since formalized in its Master Plan Part IV, which states that “autonomous vehicles have the capacity to dramatically improve the affordability, availability and safety of transportation while reducing pollution, particularly in our increasingly dense global cities.” The same document is where Tesla lays out its “sustainable abundance” framing for Cybercab and Optimus alike, describing the two as the hardware behind an AI driven push to cut the cost of transportation and labor at scale. Whether Cybercab actually operates as a robotaxi anywhere in Asia remains an open question, considering China has already pushed an autonomous ride-hailing market that’s run on homegrown players like Baidu’s Apollo Go and Pony AI. For now, the four city tour reads as a marketing push timed to Austin’s momentum.Advertisement - Cybercab will be on display in Asia this month! Come experience the future of autonomy in Hong Kong, Tokyo, Beijing & Shanghai. pic.twitter.com/wGmmEastfX — Tesla Asia (@Tesla_Asia) September 4, 2026 — Tesla Asia (@Tesla_Asia) September 4, 2026 The timing lines up with Tesla’s only prior Cybercab appearance in the region, a booth at the China International Import Expo in Shanghai last November, which Teslarati covered at the time. At that event, Tesla’s regional general manager for Shanghai framed the car as evidence of the company’s broader mission, a message Tesla has since formalized in its Master Plan Part IV, which states that “autonomous vehicles have the capacity to dramatically improve the affordability, availability and safety of transportation while reducing pollution, particularly in our increasingly dense global cities.” The same document is where Tesla lays out its “sustainable abundance” framing for Cybercab and Optimus alike, describing the two as the hardware behind an AI driven push to cut the cost of transportation and labor at scale. Whether Cybercab actually operates as a robotaxi anywhere in Asia remains an open question, considering China has already pushed an autonomous ride-hailing market that’s run on homegrown players like Baidu’s Apollo Go and Pony AI. For now, the four city tour reads as a marketing push timed to Austin’s momentum.Advertisement - The timing lines up with Tesla’s only prior Cybercab appearance in the region, a booth at the China International Import Expo in Shanghai last November, which Teslarati covered at the time. At that event, Tesla’s regional general manager for Shanghai framed the car as evidence of the company’s broader mission, a message Tesla has since formalized in its Master Plan Part IV, which states that “autonomous vehicles have the capacity to dramatically improve the affordability, availability and safety of transportation while reducing pollution, particularly in our increasingly dense global cities.” The same document is where Tesla lays out its “sustainable abundance” framing for Cybercab and Optimus alike, describing the two as the hardware behind an AI driven push to cut the cost of transportation and labor at scale. Whether Cybercab actually operates as a robotaxi anywhere in Asia remains an open question, considering China has already pushed an autonomous ride-hailing market that’s run on homegrown players like Baidu’s Apollo Go and Pony AI. For now, the four city tour reads as a marketing push timed to Austin’s momentum.Advertisement - Whether Cybercab actually operates as a robotaxi anywhere in Asia remains an open question, considering China has already pushed an autonomous ride-hailing market that’s run on homegrown players like Baidu’s Apollo Go and Pony AI. For now, the four city tour reads as a marketing push timed to Austin’s momentum.Advertisement - Tesla launched a $350 tailgate track and a $985 lockable Bed Gear Box for Cybertruck. Published on By Tesla’s Cybertruck team added two more items to the Tesla Shop, targeting job site crews and owners who use the truck bed for actual work rather than just showing it off. The official Cybertruck X account posted the Tailgate Utility Track and the Bed Gear Box within minutes of each other, part of a five item batch that also included a reflective jacket, a spray paint hat and an updated reflective tee. The Tailgate Utility Track runs $350 and turns the folded down tailgate into another mounting surface. It’s a single aluminum track with a T-slot for sliding accessories and two L-track attachment points, plus two load stops included in the box. The pitch is straightforward: strap down oversized cargo, like lumber or a cooler, that hangs off the back of the bed without it sliding out mid-drive. It bolts onto the existing tailgate and works on every Cybertruck trim. Tailgate Utility Trackhttps://t.co/PCAYXlFBvS pic.twitter.com/sCV2NVxf1W — Cybertruck (@cybertruck) September 3, 2026 The Bed Gear Box costs $985 and is a different kind of accessory. It’s a lockable aluminum storage box, 55.78 inches long, 19.8 inches wide and 7.79 inches tall, that mounts to the bed’s L-track rails and comes with two internal bins for smaller items. According to Tesla, at just over 57 pounds empty, it’s meant to stay in place rather than come in and out with each trip, giving owners a factory-fit alternative to loose totes for tools, recovery gear or emergency supplies. Tesla’s listing notes that Long Range and Dual Motor AWD Cybertrucks need the L-Tracks accessory installed separately before the Gear Box will mount, since L-tracks come standard only on certain configurations. Tesla Cybertruck bed gear box accessory Both accessories lean on the idea Tesla has been building toward since Elon Musk first described the Cybertruck’s third-party attachment strategy at the 2023 shareholder meeting, when he said the truck would ship with mounting points so outside companies, and Tesla itself, could keep adding gear without redesigning the bed. That’s the same L-track backbone underneath the tailgate shield and jumpseats Tesla launched last year, and the off-road armor package that arrived through the same X account in 2025. Owners looking to round out the rest of the L-track ecosystem, cargo dividers, MOLLE panels, bed racks and similar gear, can find a wider range of options through our Cybertruck accessories collection.Advertisement - The Tailgate Utility Track runs $350 and turns the folded down tailgate into another mounting surface. It’s a single aluminum track with a T-slot for sliding accessories and two L-track attachment points, plus two load stops included in the box. The pitch is straightforward: strap down oversized cargo, like lumber or a cooler, that hangs off the back of the bed without it sliding out mid-drive. It bolts onto the existing tailgate and works on every Cybertruck trim. Tailgate Utility Trackhttps://t.co/PCAYXlFBvS pic.twitter.com/sCV2NVxf1W — Cybertruck (@cybertruck) September 3, 2026 The Bed Gear Box costs $985 and is a different kind of accessory. It’s a lockable aluminum storage box, 55.78 inches long, 19.8 inches wide and 7.79 inches tall, that mounts to the bed’s L-track rails and comes with two internal bins for smaller items. According to Tesla, at just over 57 pounds empty, it’s meant to stay in place rather than come in and out with each trip, giving owners a factory-fit alternative to loose totes for tools, recovery gear or emergency supplies. Tesla’s listing notes that Long Range and Dual Motor AWD Cybertrucks need the L-Tracks accessory installed separately before the Gear Box will mount, since L-tracks come standard only on certain configurations. Tesla Cybertruck bed gear box accessory Both accessories lean on the idea Tesla has been building toward since Elon Musk first described the Cybertruck’s third-party attachment strategy at the 2023 shareholder meeting, when he said the truck would ship with mounting points so outside companies, and Tesla itself, could keep adding gear without redesigning the bed. That’s the same L-track backbone underneath the tailgate shield and jumpseats Tesla launched last year, and the off-road armor package that arrived through the same X account in 2025. Owners looking to round out the rest of the L-track ecosystem, cargo dividers, MOLLE panels, bed racks and similar gear, can find a wider range of options through our Cybertruck accessories collection.Advertisement - Tailgate Utility Trackhttps://t.co/PCAYXlFBvS pic.twitter.com/sCV2NVxf1W — Cybertruck (@cybertruck) September 3, 2026 — Cybertruck (@cybertruck) September 3, 2026 The Bed Gear Box costs $985 and is a different kind of accessory. It’s a lockable aluminum storage box, 55.78 inches long, 19.8 inches wide and 7.79 inches tall, that mounts to the bed’s L-track rails and comes with two internal bins for smaller items. According to Tesla, at just over 57 pounds empty, it’s meant to stay in place rather than come in and out with each trip, giving owners a factory-fit alternative to loose totes for tools, recovery gear or emergency supplies. Tesla’s listing notes that Long Range and Dual Motor AWD Cybertrucks need the L-Tracks accessory installed separately before the Gear Box will mount, since L-tracks come standard only on certain configurations. Tesla Cybertruck bed gear box accessory Both accessories lean on the idea Tesla has been building toward since Elon Musk first described the Cybertruck’s third-party attachment strategy at the 2023 shareholder meeting, when he said the truck would ship with mounting points so outside companies, and Tesla itself, could keep adding gear without redesigning the bed. That’s the same L-track backbone underneath the tailgate shield and jumpseats Tesla launched last year, and the off-road armor package that arrived through the same X account in 2025. Owners looking to round out the rest of the L-track ecosystem, cargo dividers, MOLLE panels, bed racks and similar gear, can find a wider range of options through our Cybertruck accessories collection.Advertisement - The Bed Gear Box costs $985 and is a different kind of accessory. It’s a lockable aluminum storage box, 55.78 inches long, 19.8 inches wide and 7.79 inches tall, that mounts to the bed’s L-track rails and comes with two internal bins for smaller items. According to Tesla, at just over 57 pounds empty, it’s meant to stay in place rather than come in and out with each trip, giving owners a factory-fit alternative to loose totes for tools, recovery gear or emergency supplies. Tesla’s listing notes that Long Range and Dual Motor AWD Cybertrucks need the L-Tracks accessory installed separately before the Gear Box will mount, since L-tracks come standard only on certain configurations. Tesla Cybertruck bed gear box accessory Both accessories lean on the idea Tesla has been building toward since Elon Musk first described the Cybertruck’s third-party attachment strategy at the 2023 shareholder meeting, when he said the truck would ship with mounting points so outside companies, and Tesla itself, could keep adding gear without redesigning the bed. That’s the same L-track backbone underneath the tailgate shield and jumpseats Tesla launched last year, and the off-road armor package that arrived through the same X account in 2025. Owners looking to round out the rest of the L-track ecosystem, cargo dividers, MOLLE panels, bed racks and similar gear, can find a wider range of options through our Cybertruck accessories collection.Advertisement - Tesla Cybertruck bed gear box accessory Both accessories lean on the idea Tesla has been building toward since Elon Musk first described the Cybertruck’s third-party attachment strategy at the 2023 shareholder meeting, when he said the truck would ship with mounting points so outside companies, and Tesla itself, could keep adding gear without redesigning the bed. That’s the same L-track backbone underneath the tailgate shield and jumpseats Tesla launched last year, and the off-road armor package that arrived through the same X account in 2025. Owners looking to round out the rest of the L-track ecosystem, cargo dividers, MOLLE panels, bed racks and similar gear, can find a wider range of options through our Cybertruck accessories collection.Advertisement - Owners looking to round out the rest of the L-track ecosystem, cargo dividers, MOLLE panels, bed racks and similar gear, can find a wider range of options through our Cybertruck accessories collection.Advertisement - Published on By Tesla Cybercab rides are officially open to the public in Austin, Texas, as the company confirmed on Thursday following its launch event that the two-seater would be available in the company’s Robotaxi fleet. Cybercab is Tesla’s first vehicle completely void of any manual controls. It has no steering wheel and no pedals, and it will utilize Tesla’s Full Self-Driving fleet to operate. The first rides have already happened, as those at the event were able to hail a Cybercab to any location within the company’s geofence in Austin. Tesla’s $25K car is the Cybercab with no steering wheel or pedals The addition of Cybercab to the public Robotaxi fleet is a major statement in Tesla’s trek to launch fully autonomous driving. For years, critics have complained about the need for drivers to continuously supervise the vehicle. With Cybercab, there are no manual controls in the cockpit other than to control the seat, the center screen, and the climate. The vehicle is fully geared toward being a living room on wheels in a sense: equipped with Starlink V5 satellites, CEO Elon Musk said the vehicle would enable 4K live video, gaming, and other entertainment options during travel. Cybercab is basically a super comfortable lounge on wheels with a great TV and epic sound https://t.co/X1aWKoJdJV — Elon Musk (@elonmusk) September 3, 2026Advertisement - Musk noted that Cybercab is “designed and built for maximally efficient autonomous operation.” Tesla continues to push the envelope on autonomy, and over the next several months, the company could start selling Cybercab units to the public. The company opened up a public interest form on its website to gauge demand, and many have already submitted requests to purchase a fleet of Cybercab units for their own personal ride-hailing side hustle. Tesla hints its already prepping for Cybercab fleet orders The launch of Cybercab in this area marks a major accomplishment for Tesla, as it also announced that it has reached 1 million unsupervised autonomous miles since launching driverless rides on the Robotaxi fleet.Advertisement - Things are moving along at a fine pace, and although we have waited for this for some time, the day has finally come when Tesla is offering self-driving rides of some kind to the public. Cybercab is Tesla’s first vehicle completely void of any manual controls. It has no steering wheel and no pedals, and it will utilize Tesla’s Full Self-Driving fleet to operate. The first rides have already happened, as those at the event were able to hail a Cybercab to any location within the company’s geofence in Austin. Tesla’s $25K car is the Cybercab with no steering wheel or pedals The addition of Cybercab to the public Robotaxi fleet is a major statement in Tesla’s trek to launch fully autonomous driving. For years, critics have complained about the need for drivers to continuously supervise the vehicle. With Cybercab, there are no manual controls in the cockpit other than to control the seat, the center screen, and the climate. The vehicle is fully geared toward being a living room on wheels in a sense: equipped with Starlink V5 satellites, CEO Elon Musk said the vehicle would enable 4K live video, gaming, and other entertainment options during travel. Cybercab is basically a super comfortable lounge on wheels with a great TV and epic sound https://t.co/X1aWKoJdJV — Elon Musk (@elonmusk) September 3, 2026Advertisement - Musk noted that Cybercab is “designed and built for maximally efficient autonomous operation.” Tesla continues to push the envelope on autonomy, and over the next several months, the company could start selling Cybercab units to the public. The company opened up a public interest form on its website to gauge demand, and many have already submitted requests to purchase a fleet of Cybercab units for their own personal ride-hailing side hustle. Tesla hints its already prepping for Cybercab fleet orders The launch of Cybercab in this area marks a major accomplishment for Tesla, as it also announced that it has reached 1 million unsupervised autonomous miles since launching driverless rides on the Robotaxi fleet.Advertisement - Things are moving along at a fine pace, and although we have waited for this for some time, the day has finally come when Tesla is offering self-driving rides of some kind to the public. Tesla’s $25K car is the Cybercab with no steering wheel or pedals The addition of Cybercab to the public Robotaxi fleet is a major statement in Tesla’s trek to launch fully autonomous driving. For years, critics have complained about the need for drivers to continuously supervise the vehicle. With Cybercab, there are no manual controls in the cockpit other than to control the seat, the center screen, and the climate. The vehicle is fully geared toward being a living room on wheels in a sense: equipped with Starlink V5 satellites, CEO Elon Musk said the vehicle would enable 4K live video, gaming, and other entertainment options during travel. Cybercab is basically a super comfortable lounge on wheels with a great TV and epic sound https://t.co/X1aWKoJdJV — Elon Musk (@elonmusk) September 3, 2026Advertisement - Musk noted that Cybercab is “designed and built for maximally efficient autonomous operation.” Tesla continues to push the envelope on autonomy, and over the next several months, the company could start selling Cybercab units to the public. The company opened up a public interest form on its website to gauge demand, and many have already submitted requests to purchase a fleet of Cybercab units for their own personal ride-hailing side hustle. Tesla hints its already prepping for Cybercab fleet orders The launch of Cybercab in this area marks a major accomplishment for Tesla, as it also announced that it has reached 1 million unsupervised autonomous miles since launching driverless rides on the Robotaxi fleet.Advertisement - Things are moving along at a fine pace, and although we have waited for this for some time, the day has finally come when Tesla is offering self-driving rides of some kind to the public. The addition of Cybercab to the public Robotaxi fleet is a major statement in Tesla’s trek to launch fully autonomous driving. For years, critics have complained about the need for drivers to continuously supervise the vehicle. With Cybercab, there are no manual controls in the cockpit other than to control the seat, the center screen, and the climate. The vehicle is fully geared toward being a living room on wheels in a sense: equipped with Starlink V5 satellites, CEO Elon Musk said the vehicle would enable 4K live video, gaming, and other entertainment options during travel. Cybercab is basically a super comfortable lounge on wheels with a great TV and epic sound https://t.co/X1aWKoJdJV — Elon Musk (@elonmusk) September 3, 2026Advertisement - Musk noted that Cybercab is “designed and built for maximally efficient autonomous operation.” Tesla continues to push the envelope on autonomy, and over the next several months, the company could start selling Cybercab units to the public. The company opened up a public interest form on its website to gauge demand, and many have already submitted requests to purchase a fleet of Cybercab units for their own personal ride-hailing side hustle. Tesla hints its already prepping for Cybercab fleet orders The launch of Cybercab in this area marks a major accomplishment for Tesla, as it also announced that it has reached 1 million unsupervised autonomous miles since launching driverless rides on the Robotaxi fleet.Advertisement - Things are moving along at a fine pace, and although we have waited for this for some time, the day has finally come when Tesla is offering self-driving rides of some kind to the public. With Cybercab, there are no manual controls in the cockpit other than to control the seat, the center screen, and the climate. The vehicle is fully geared toward being a living room on wheels in a sense: equipped with Starlink V5 satellites, CEO Elon Musk said the vehicle would enable 4K live video, gaming, and other entertainment options during travel. Cybercab is basically a super comfortable lounge on wheels with a great TV and epic sound https://t.co/X1aWKoJdJV — Elon Musk (@elonmusk) September 3, 2026Advertisement - Musk noted that Cybercab is “designed and built for maximally efficient autonomous operation.” Tesla continues to push the envelope on autonomy, and over the next several months, the company could start selling Cybercab units to the public. The company opened up a public interest form on its website to gauge demand, and many have already submitted requests to purchase a fleet of Cybercab units for their own personal ride-hailing side hustle. Tesla hints its already prepping for Cybercab fleet orders The launch of Cybercab in this area marks a major accomplishment for Tesla, as it also announced that it has reached 1 million unsupervised autonomous miles since launching driverless rides on the Robotaxi fleet.Advertisement - Things are moving along at a fine pace, and although we have waited for this for some time, the day has finally come when Tesla is offering self-driving rides of some kind to the public. Cybercab is basically a super comfortable lounge on wheels with a great TV and epic sound https://t.co/X1aWKoJdJV — Elon Musk (@elonmusk) September 3, 2026Advertisement - — Elon Musk (@elonmusk) September 3, 2026Advertisement - Musk noted that Cybercab is “designed and built for maximally efficient autonomous operation.” Tesla continues to push the envelope on autonomy, and over the next several months, the company could start selling Cybercab units to the public. The company opened up a public interest form on its website to gauge demand, and many have already submitted requests to purchase a fleet of Cybercab units for their own personal ride-hailing side hustle. Tesla hints its already prepping for Cybercab fleet orders The launch of Cybercab in this area marks a major accomplishment for Tesla, as it also announced that it has reached 1 million unsupervised autonomous miles since launching driverless rides on the Robotaxi fleet.Advertisement - Things are moving along at a fine pace, and although we have waited for this for some time, the day has finally come when Tesla is offering self-driving rides of some kind to the public. Musk noted that Cybercab is “designed and built for maximally efficient autonomous operation.” Tesla continues to push the envelope on autonomy, and over the next several months, the company could start selling Cybercab units to the public. The company opened up a public interest form on its website to gauge demand, and many have already submitted requests to purchase a fleet of Cybercab units for their own personal ride-hailing side hustle. Tesla hints its already prepping for Cybercab fleet orders The launch of Cybercab in this area marks a major accomplishment for Tesla, as it also announced that it has reached 1 million unsupervised autonomous miles since launching driverless rides on the Robotaxi fleet.Advertisement - Things are moving along at a fine pace, and although we have waited for this for some time, the day has finally come when Tesla is offering self-driving rides of some kind to the public. Tesla continues to push the envelope on autonomy, and over the next several months, the company could start selling Cybercab units to the public. The company opened up a public interest form on its website to gauge demand, and many have already submitted requests to purchase a fleet of Cybercab units for their own personal ride-hailing side hustle. Tesla hints its already prepping for Cybercab fleet orders The launch of Cybercab in this area marks a major accomplishment for Tesla, as it also announced that it has reached 1 million unsupervised autonomous miles since launching driverless rides on the Robotaxi fleet.Advertisement - Things are moving along at a fine pace, and although we have waited for this for some time, the day has finally come when Tesla is offering self-driving rides of some kind to the public. The company opened up a public interest form on its website to gauge demand, and many have already submitted requests to purchase a fleet of Cybercab units for their own personal ride-hailing side hustle. Tesla hints its already prepping for Cybercab fleet orders The launch of Cybercab in this area marks a major accomplishment for Tesla, as it also announced that it has reached 1 million unsupervised autonomous miles since launching driverless rides on the Robotaxi fleet.Advertisement - Things are moving along at a fine pace, and although we have waited for this for some time, the day has finally come when Tesla is offering self-driving rides of some kind to the public. Tesla hints its already prepping for Cybercab fleet orders The launch of Cybercab in this area marks a major accomplishment for Tesla, as it also announced that it has reached 1 million unsupervised autonomous miles since launching driverless rides on the Robotaxi fleet.Advertisement - Things are moving along at a fine pace, and although we have waited for this for some time, the day has finally come when Tesla is offering self-driving rides of some kind to the public. The launch of Cybercab in this area marks a major accomplishment for Tesla, as it also announced that it has reached 1 million unsupervised autonomous miles since launching driverless rides on the Robotaxi fleet.Advertisement - Things are moving along at a fine pace, and although we have waited for this for some time, the day has finally come when Tesla is offering self-driving rides of some kind to the public. Things are moving along at a fine pace, and although we have waited for this for some time, the day has finally come when Tesla is offering self-driving rides of some kind to the public. How to give your Tesla a Custom Lovk Sound! Easy tutorial!! #tesla #teslatok #teslalocksound Copyright © TESLARATI. All rights reserved.

O Modo Escuro é sustentável e inteligente! Com o apoio do Poupança no Minuto, onde a poupança vai muito além do ecrã. 6 Fotos © Tesla Tech Tesla A Tesla aproveitou o seu evento anual AI Day para proporcionar um vislumbre do protótipo do robot humanóide que se encontra a desenvolver - o Optimus ou Tesla Bot, como também é conhecido. Apresentado como um concept no AI Day de 2021, o Optimus teve direito a uma demonstração mais completa na edição deste ano que teve a duração de sete minutos. Durante este período, o protótipo do Optimus foi capaz de andar e acenar para os presentes na apresentação. No entanto, a Tesla partilhou um vídeo onde é possível ver o Optimus a realizar algumas tarefas, nomeadamente dar água a plantas, carregar caixas e realizar outras funções para os quais foi desenvolvido. Pode ver abaixo a apresentação do Optimus e, na galeria acima, algumas imagens do modelo atual deste robot da Tesla. Leia Também: Um dos fundadores da Airbnb juntou-se à administração da Tesla Recomendados para si Patrocinado por hôma Decoração de outono: Vista a casa com charme em tons castanhos e dourados Patrocinado por Decathlon Vitalsport: Setembro traz (de volta) a festa de desporto da Decathlon! Patrocinado por Poupança No Minuto 5 erros que podem estar a encarecer o seu crédito habitação Newsletter Receba os principais destaques todos os dias no seu email. Mais lidas Newsletter Receba os principais destaques todos os dias no seu email. Menu App eleita produto do ano 2026 pela segunda vez Seja sempre o primeiro a saber. Descarregue a nossa App gratuita

Tesla’s Optimus program holds an estimated 1,000 to 1,200 humanoid robot units deployed across Fremont and Giga Texas as of mid-2026, yet the company reports zero external sales, publishes no uptime figures, and Musk himself described those units on the Q4 2025 earnings call as “primarily for learning and data collection rather than performing productive tasks.” Tesla converted its Fremont Model S/X line to Optimus assembly in mid-2026, with initial units going to internal training rather than commercial customers. Tesla ended Model S/X production at Fremont in early May 2026 and spent four months converting that line for Optimus assembly. As of the Q2 2026 earnings call on August 5, 2026, Musk stated production was starting “soon,” targeted for late July or August 2026. No unit count was disclosed for Q2. Early Fremont builds go to the “Optimus Academy,” an internal program where robots learn simple factory skills before advancing to productive tasks. The production ramp also faces a material supply constraint. China imposed April 2025 export controls on seven medium and heavy rare earth elements, including terbium and dysprosium used in NdFeB servo magnets. The IEA reports China controls 94% of global sintered permanent magnet production; after those April 2025 controls, European rare earth magnet prices reached up to six times Chinese levels. Each humanoid robot requires approximately 3.5 kg of NdFeB magnets across 40-plus servo actuators. China suspended a broader second set of controls through November 2026, but the underlying supply concentration is unresolved. Confirmed tasks include battery cell sorting, parts handling, kitting, quality inspection, and pick-and-place, but Tesla has published no throughput or error rate data. The verified internal task list for Optimus covers sorting 4680 battery cells, moving parts between stations, kitting, quality inspection, and pick-and-place. Tesla has not published cycle times, error rates, or throughput for any of these tasks. On the Q2 2026 call, Musk stated: “This is going to be the hardest product to scale manufacturing that we’ve ever made at Tesla because everything on the robot is new.” With over 10,000 unique components, he noted the ramp “will move as fast as the least lucky, slowest, dumbest part in the entire 10,000.” Figure AI and Agility Robotics have paying external customers with documented operating hours; Optimus has zero external deployments and no verified performance data. As of mid-2026, no humanoid robot from any manufacturer has been deployed above the low hundreds of units in a sustained commercial environment, per a July 2026 Technology.org analysis. Within that constrained field, Figure AI and Agility Robotics hold a clear advantage: data from paying external customers. Figure AI’s BMW Spartanburg deployment runs ten-hour shifts loading sheet-metal parts into welding fixtures, with 90,000-plus parts and above 99% placement accuracy. Agility’s RoboFab facility has stated production capacity for 10,000 Digit units annually. Current manufacturing cost runs $50,000 to $100,000 per unit, with AI5 chip integration delayed until mid-2027 and rare earth supply risks unresolved. Current manufacturing cost is estimated at $50,000 to $100,000 per unit. Musk stated at Davos in January 2026 that the consumer target price is “under $20,000,” with a long-term mass-production price target of $20,000 to $30,000. Industry estimates based on Bureau of Labor Statistics data put fully loaded US manufacturing labor at $95,000 to $120,000 per year. Optimus will not receive Tesla’s AI5 chip until volume production is reached, projected around mid-2027. Wolfe Research considers meaningful external revenue unlikely before late 2027. Goldman Sachs projects a $38 billion humanoid robot market by 2035; Bank of America projects 90,000 units shipped industry-wide in 2026, scaling to 1.2 million by 2030. AI tools were used to assist with research and data compilation for this article. Edited and verified by the IIoT World editorial team. Axis Intelligence Research estimates 1,000 to 1,200 Optimus units at Fremont and Giga Texas as of mid-2026. Tesla has not confirmed this. On the Q4 2025 earnings call, Musk stated the units are “primarily for learning and data collection rather than performing productive tasks.” Confirmed tasks include sorting 4680 battery cells, moving parts between production stations, kitting, quality inspection, and pick-and-place. Tesla has not published cycle times, error rates, or throughput figures for any of these tasks. Figure AI logged 1,250-plus hours at BMW Spartanburg, with 90,000-plus parts at above 99% placement accuracy. Agility Robotics accumulated 65,000-plus hours across nine commercial facilities. Optimus has zero external customers and no independently verified performance data as of August 2026. Current manufacturing cost is estimated at $50,000 to $100,000 per unit. Musk stated at Davos in January 2026 that the consumer target price is “under $20,000,” with a long-term mass-production price target of $20,000 to $30,000. Industry estimates based on Bureau of Labor Statistics data put fully loaded US manufacturing labor at $95,000 to $120,000 per year. Early builds go to the Optimus Academy for internal training, not external customers. Wolfe Research considers meaningful revenue unlikely before late 2027. AI5 chip integration for Optimus is projected around mid-2027. Fifty-seven percent of manufacturing executives report that inadequate data quality hampers AI use-case development, according to MIT Technology Review. Meanwhile,… Manufacturers generate enormous volumes of data from machines, equipment, and enterprise systems, yet 84% of companies report that data access… According to BMW Group, the automaker runs virtual replicas of more than 30 production sites where employees test layout changes… © 2017-2026 IIoT World. All articles submitted by our contributors do not constitute the views, endorsements or opinions of IIoT-World.com. Sep 09–10 / Virtual / Free LIVE PANELS on agentic AI, edge deployment, digital twins, and measurable ROI on the plant floor. We are running our annual survey. Your experience becomes the benchmark the industry cites next year, and it is yours, free.

Die besten Autos Einzeltest Dauertest Vergleichstest Gebrauchtwagentest Kaufberatung Produkttests Reifentests Produktvergleich Auto-News New Mobility Motorsport Panorama Verkehr Deals Firmenwagen Motorrad Nutzfahrzeuge Tuning Wohnmobile Wohnwagen Kommentare Messen SUV Auto-Abo Autokauf Autoverkauf Autofinanzierung Neuwagenkauf Leasing Autopflege Autoteile Gebrauchtwagen Führerschein Kfz-Versicherung Mobilität Recht Reifen Reparatur & Technik Sicherheit Fahrbericht Neuvorstellung Motorsounds Ratgeber Test Die besten Autohändler Die besten Reisemobilhändler Die besten Werkstätten Die besten Waschanlagen Die besten Auto & Motorrad Shops AUTO BILD CLUB Allradautos des Jahres Die besten Importautos Die besten Marken aller Klassen Die wertstabilsten Autos Firmenwagen Award Goldener Klassiker Goldenes Lenkrad Goldenes Reisemobil Motorsport Award Sportscars des Jahres 40 Jahre AUTO BILD AUTO BILD e-Motion Days AUTO BILD Storys AITO Intelligenter Luxus Euromaster zahlt Ihre Rechnung Mazda6e Leser-Dauertest Hankook Leserreise Hot Wheels Legends Tour Leapmotor C10 REEV Toyota Stories Service-Links

by Alimat Aliyeva A unique “kindergarten” for robots has been opened in China, where researchers are testing a new approach to machine learning. Instead of teaching robots every action in advance, scientists want them to learn independently through experiments, mistakes, and repeated attempts. The training center was created by the Chinese company Tashan Technology. Researchers believe that this method could allow robots to continue adapting even after they are deployed in the real world, including in situations they were never specifically trained for. For example, if a robot crashes into a wall during training, it can analyze the failure, change its route, and try again. In this way, mistakes become useful learning data rather than simply being treated as failures. Wang Peng, a researcher at the Beijing Academy of Social Sciences, said that independent exploration of the environment could help robots move beyond narrow training for specific tasks. Instead, they may gradually develop a more fundamental understanding of the physical world and how objects around them behave. However, this type of learning requires much more time. Robots may need to make hundreds or even thousands of attempts before they learn how to deal with a new situation. Still, researchers believe that the ability to make mistakes and learn from them could become an important part of the future of robotics. In the long term, scientists also hope that robots will be able to share their experiences with one another. If one robot learns how to overcome a particular obstacle or complete a new task, other robots could potentially use that knowledge instead of learning everything from scratch. Interestingly, this approach is similar to the way young children learn. They do not simply follow instructions — they explore, make mistakes, and gradually understand how the world works. Perhaps the idea of a “kindergarten” is therefore more than just a funny name: it reflects a completely different way of teaching machines. Here we are to serve you with news right now. It does not cost much, but worth your attention. Choose to support open, independent, quality journalism and subscribe on a monthly basis. By subscribing to our online newspaper, you can have full digital access to all news, analysis, and much more. You can also follow AzerNEWS on Twitter @AzerNewsAz or Facebook @AzerNewsNewspaper Thank you! © Azernews.az 2026

The flashy company, which recently completed a blockbuster IPO, appears to be leading the pack of humanoid robot makers. Image Credit Unitree via X Share Increasingly, companies are building humanoid robots that perform impressive athletic feats to mark the field’s progress. Now, Chinese robotics company Unitree says its new "Superman" robot can run 12.66 meters per second, faster than Usain Bolt's top recorded speed. Getting a humanoid robot to run at all requires split-second control and has been a significant engineering challenge occupying roboticists for decades. That’s why sprinting, as well as jumping, have become popular targets for robotics companies keen to demonstrate their technology’s prowess. Unitree's latest demonstration pushes the boundaries by not only outrunning the fastest human ever, but also jumping around 6 feet 7 inches into the air from a standing start, a full foot more than the human record. “This new machine has only been in development for a little over three months, with significant room for further improvement in the coming months,” Unitree said in an X post that accompanied a video of the accomplishments. The records have not been externally verified, and the sprinting speed was a peak reading taken over a shorter stretch rather than a full 100 meters like Bolt’s record. The robot's legs are also only 2 feet 9 inches long, according to Unitree, which results in an ungainly, arm-waving gait while running. The effort is nonetheless impressive and adds to Unitree’s growing reputation as the company leading the pack of humanoid robot developers. And the timing of the announcement was no accident, coming just days before Unitree's stock market debut and shortly before the World Humanoid Robot Games, which opened on August 22. The company’s Shanghai IPO was a blockbuster, recording an initial 629 percent gain on the company’s first day of trading. It was briefly valued at around $66 billion before closing at a more modest $51 billion. However, some analysts have cautioned the excitement around the company’s technology may be getting ahead of market realities. “The IPO is expensive, and the investment ​risk is already ​quite high,” ⁠Wang Zhuo, partner of Shanghai Zhuozhu Investment Management, told Reuters. “Unitree generates much of its sales from research and demonstrations, but ​wider application is still far away.” Sign up to receive top stories about groundbreaking technologies and visionary thinkers from SingularityHub. But the company holds a dominant grip on the emerging humanoid market that may justify some of the hype. Chinese firms control roughly 90 percent of the global humanoid robot market, with Unitree alone shipping 5,500 of the 13,000 to 18,000 humanoids sold worldwide in 2025, the most of any manufacturer. In contrast, US humanoid champions Figure AI, Agility Robotics, and Tesla each shipped around 150 units. China’s success is down to “a combination of policy support, public investment, mature supply chain, and advancements made in AI software and hardware,” Lian Jye Su, a tech analyst at consultancy firm Omdia, told Rest of World. This is leading to an increasingly combative response from the US. On July 29 the Federal Communications Commission banned new imports of foreign-made humanoid and quadruped robots. The move was framed as a matter of national security, though it has also been seen as an attempt to give domestic developers a leg up. Beijing predictably objected, with foreign ministry spokesperson Mao Ning telling a press conference that “protectionism does not make the US more competitive, and it will only hurt the interests of US companies and consumers.” Given the rapid progress made by companies like Unitree, it seems likely it’s going to take more than trade barriers for the US to catch up. In the meantime, we might see more human athletic records fall to China’s leading humanoid developers. Related Articles What we’re reading Sign up to receive top stories about groundbreaking technologies and visionary thinkers from SingularityHub. SingularityHub chronicles the technological frontier with coverage of the breakthroughs, players, and issues shaping the future.

Across the Glasgow City region, there’s never a day that goes by without an exciting new development within the world of Science and Innovation. Our weekly newsletter will keep you up-to-date with the latest news and exciting events happening across the region in an easy-to-digest format. Sign up below! You can unsubscribe at any time by clicking the link in the footer of our emails. For information about our privacy practices, visit our privacy policy page. Giving robot companions animal-like or anime-style faces and voices could maximise their emotional appeal to users, according to new research which could help guide the growing field of social robotics. Social robots can provide companionship and comfort to their owners by mimicking human or petlike behaviours. Although the robots are a fast-growing sector of the tech industry, surprisingly little academic research has explored how their designs can be optimised to make the robots as expressive and emotionally engaging as possible. In a new paper set to be presented at a conference this week, researchers from the University of Glasgow show how they have been working to close that research gap. With the help of a group of volunteers, they used augmented reality technology to test dozens of different combinations of faces and voices overlaid on the body of an existing social robot called a Qoobo, which has a soft, cushion-like body and a swinging tail. Giving robot companions animal-like or anime-style faces and voices could maximise their emotional appeal to users, according to new research which could help guide the growing field of social robotics. Social robots can provide companionship and comfort to their owners by mimicking human or petlike behaviours. Although the robots are a fast-growing sector of the tech industry, surprisingly little academic research has explored how their designs can be optimised to make the robots as expressive and emotionally engaging as possible. In a new paper set to be presented at a conference this week, researchers from the University of Glasgow show how they have been working to close that research gap. With the help of a group of volunteers, they used augmented reality technology to test dozens of different combinations of faces and voices overlaid on the body of an existing social robot called a Qoobo, which has a soft, cushion-like body and a swinging tail. Dr Shaun Macdonald and Josh Yip of the University of Glasgow’s School of Computing Science used this rapid prototyping approach to explore how users felt about the virtual robots, and how effectively they were able to ‘read’ the robot’s emotional states to make them feel more like real creatures. Dr Macdonald said: “Social robots can provide valuable experiences for a wide range of people. Older people have reported that they can help mitigate feelings of loneliness or social isolation, for example, and people living in spaces where they can’t keep a real pet have kindled some of that sense of companionship with what we call zoomorphic robots, which mimic animals’ appearances and behaviour. “However, the aesthetic design of zoomorphic social robots so far has largely been the result of educated guesswork, with very little research available on what makes them appealing to users. “What we’ve done here for the first time here is take a more systematic approach to designing the appearance of a zoomorphic robot by harnessing augmented reality to enable rapid, side-by-side design tests. In the future, these insights could help create a new generation of user-customisable robots which can tailor their appearances to better suit their owners’ preferences.” The research builds on Dr Macdonald’s previous research in advancing the design of social robots. His work includes the development of a software system called Augmenting Zoomorphic Robotics with Affect, or AZRA, which can overlay augmented-reality sounds and animations over the bodies of real-world social robots to prototype new appearances and functionalities. In this research, the team began by creating a ‘mood board’ of appealing images of animals and robots drawn from real life and from popular culture to help them design a series of prototype faces for virtual robots. They ended up with five prototype faces comprised of animal-like, robot-like, emoji-style, anime-inspired designs, with another comprised simply of eyes. They chose five distinct vocalisations to enable the prototypes to express themselves to the volunteers: recordings of cat noises, human-like vocal sounds, abstract electronic noises, music and samples of ‘animalese’, the nonsense sounds used by the characters in the popular Nintendo game series Animal Crossing. Then, 24 volunteers were given AR headsets and asked to rank different combinations of faces and voices overlaid on Qoobo for how clearly they expressed an emotion and how much empathy they felt towards the prototype robot, as well as how appropriate the combination of factors felt to them. The results revealed that combinations of animal-like faces and vocalisations matched Qoobo’s animal-like body were well-liked by the volunteers, who found them more engaging and immersive. While that result might be unsurprising, the team also found that anime-like faces were just as highly-ranked despite not being based on real life. Josh Yip said: “Anime-like faces are more expressive than animal faces could ever really be, giving the robot abilities to express themselves beyond those of a normal pet but in a way that didn’t feel uncanny. It still felt like it could be a pet-like creature to our study participants. “We also found that faces consistently outperformed sounds at conveying emotion and creating connections. Interestingly, the study participants found that the human sounds were easier to read emotionally but were liked less than animal sounds, even though they found them harder to clearly identify with specific emotions.” The team were surprised to find that, although the study participants insisted that they felt that being able to correctly read the robot’s emotions was very important to them, the data showed that readability had little influence on the designs they preferred. Along similar lines, ‘animalese’ sounds ranked highly with users despite being unintelligible, with some users feeling it suggested the robot had a greater sense of intelligence than other sounds. The study’s participants reactions to the prototypes were also affected by whether they owned a real animal pet themselves. Pet owners were more likely to treat the animal-like prototypes as real pets, but to be more likely to treat robot-like faces as machines. Dr Macdonald added: “Our research suggests that although there is no single face or voice that has universal appeal, the combination of the two can have a big effect on how well users feel that robots are able to communicate emotions. “We hope our findings will help inform the design of future generations of zoomorphic social robot technology, but they could also extend the lifespan of existing devices by enabling users to overlay new faces, voices and functionality on them using augmented reality.” The team’s paper, titled ‘Anime, Animal or Animal Crossing? Comparing Aesthetic Styles for Zoomorphic Robot Faces and Sounds using Augmented Reality’, will be presented at the ACM International Conference on Mobile Human-Computer Interaction conference in Swansea, Wales on Thursday 3 September. The project was supported by funding from the University of Glasgow’s EPSRC Impact Acceleration Account and the Engineering and Physical Sciences Research Council (EPSRC). T: 0141 420 5010 E: info@glasgowcityofscienceandinnovation.com T: 0141 420 5010 E: info@glasgowcityofscienceandinnovation.com © Copyright 2026 Glasgow City of Science and Innovation Site by Maguires

ByLance Eliot, Contributor. The FCC has announced a ban on foreign-produced AI humanoid robots, effective July 27, 2026, citing national security concerns. The agency fears these devices act as "Trojan Horses," collecting sensitive personal and corporate data through high-fidelity sensors and transmitting it to foreign entities. This data could be leveraged for surveillance or intelligence, and robots could be remotely commandeered. While the ban targets all foreign manufacturers, it significantly impacts China, which holds a large market share. A secondary goal is to bolster U.S. domestic manufacturing. However, the article warns that even American-made robots still pose privacy risks, urging consumers to remain vigilant about these increasingly common devices in homes and workplaces. In today’s column, I examine the newly announced ban by the U.S. on various imports of AI humanoid robots from foreign entities. The ban is under the auspices of the FCC. I provide an analysis of the key rationale and stipulations that the federal agency announced in its official proclamation of July 27, 2026. The overall gist is that people are increasingly buying and using humanoid robots in their homes and in the workplace, which turns out to be a plus but also has a sour underbelly involved too. These AI-powered roaming devices can collect all sorts of private information and readily transmit that data to the maker of the robots. A foreign entity could make use of that data in unsavory ways. Ergo, the U.S. is going to ban their importation in an effort to prevent a massive-scale adoption of Trojan Horses. There are other equally notable reasons associated with the basis for the ban. Let’s talk about it. This analysis of AI breakthroughs is part of my ongoing Forbes column coverage on the latest in AI, including identifying and explaining various impactful AI complexities (see the link here). I’ve extensively been covering the advances in humanoid robots, especially as they are being coupled with the latest generative AI and large language models (LLMs). Adding LLMs to humanoid robots is making them more useful and appealing to a wider array of users. When used in the home, there are many upsides and downsides. For example, some are tapping into the devices for mental health advice while the robots are casually wandering around the house and doing everyday chores for the homeowner. See my coverage at the link here and the link here. A humanoid robot is a type of robot that is purposely built to resemble a human in complete form and function. There is a robotic head, arms, hands, fingers, legs, body, etc. You’ve undoubtedly seen those types of robots in the many online video recordings showing them walking, jumping, grasping at objects, and so on. A tremendous amount of active research and development is taking place to perfect humanoid robots. They look rather comical right now. You watch those videos and laugh when the robot trips over a small stick lying on the ground, something that a human would seldom trip over. We scoff when a tested humanoid robot tries to grasp a coffee cup and inadvertently spills most of the liquid java. It all seems humorous and a silly pursuit. Keep in mind that we are all observing the development process while it is still taking place. At some point, those guffaws of the humanoid robots will lessen. Humanoid robots will be as smooth and graceful as humans. This will continue to be honed. Eventually, humanoid robots will be less prone to physical errors than humans are. In a sense, the physicality of a humanoid robot will be on par with humans, if not better, due to its smoothly refined mechanical and electronic properties. Do not discount the coming era of physically capable humanoid robots. This is going to have a massive impact on society. For more details on AI and humanoid robots, along with research known as physical AI, see my in-depth discussion at the link here and the link here. AI has a lot to do with humanoid robots, including these two major facets: I will briefly unpack those two facets. Humanoid robots lean into operational AI to aid in giving the robot a bunch of very fundamental or bare-bones capabilities. There is AI that does vision processing and analyzes the images being captured by the embedded cameras that are considered the eyes of the robot. The ears of the robot are recording sounds that the robot hears and feeds the audio into a specialized AI that analyzes it. All in all, a whole lot of AI is working inside the robot to enable the robot to maneuver in the real world. Beyond those core functional aspects, generative AI is being added as an integral element of humanoid robots. Here’s how that works. A customized generative AI or large language model (LLM) is placed inside the processors and data memory contained within the frame of the humanoid robot. This will provide a self-contained generative AI or LLM that doesn’t need any Internet access. We might also include Internet access so that the generative AI can reach the Internet to get additional information, but we don’t want the robot to be dependent on network accessibility. The idea is that the humanoid robot needs to have immediate access to the LLM and not be delayed if network connections are spotty. Suppose you decide to purchase the latest robot device that will clean your floors and act as a sentry in your home. The robot roams from room to room. It sweeps the tile floors and vacuums the carpet. The robot has an additional duty. It can be used while you are away to scan for any potential intruders. Nice. While you are in your home, you take little notice of the robot. It quietly moves from here to there. No big deal. You bought this particular model because it has a low profile and won’t get in your way or otherwise be obnoxious. The robot was available at a really low price, which turned out to be made by a foreign company and available for a steep discount. A few months later, you find out that the robot has been listening to your private conversations this whole time and capturing video of you. The data was transmitted to the foreign company that made the robot. They know all sorts of aspects about you, your family members, people who have visited you in your home, your habits, your comings and goings, etc. All of this is now in the hands of the foreign entity and can be used in whatever manner they choose. In a sense, the robot was a Trojan Horse. You bought it and didn’t think twice about anything other than the need to clean your floors and have a handy detector for burglars and such. The entire time, the sneaky robot was collecting information about you and giving that information to a company based in a foreign country. And, since the company is in a foreign country, your legal recourse is severely limited. Not nice. Disconcerting and highly disturbing. The FCC formally issued a publicly posted statement of determination on July 27, 2026, entitled “National Security Determination on the Threat Posed by Foreign-Produced Advanced Robotic Devices” that made these salient points (excerpts): Let’s unpack those notable points. I already pointed out that one disturbing aspect is that the AI humanoid robots can be a veritable spy. If used in a home, your personal privacy is potentially invaded. When used in the workplace, one concern is that the robots might aid in stealing corporate secrets or using insider information that could be transmitted to the foreign company for nefarious purposes. The humanoid robot is a snitch, a tattletale, a 24/7 snoop. Another concern is that these humanoid robots are likely controllable by remote access of the foreign entity. Imagine this. The robot in a workplace is given a command to go to a conference room and stay there in a quiet mode. Top executives come into the conference room to discuss corporate strategy. The robot is not noticed by anyone in the room. Meanwhile, it is livestreaming the entire discussion to the foreign company. In this case, the robot was given a remote command to sneakily be in the conference room to record the meeting. If the foreign entity wanted to do so, they could seemingly operate the robot to do destructive acts in the workplace. It is a scary proposition. Some blazing news headlines about the FCC announcement were quick to claim that this was a ban against China. Well, kind of yes, kind of no. The FCC does not explicitly name China as the target of the ban. The ban is for all foreign-produced advanced robotic devices. Naturally, this encompasses China, along with all other countries in the world. The twist is that China generally has about 85% of the humanoid robot market right now. They currently have the lion’s share of this niche. Therefore, the blanket ban of all countries will have a greater impact on China due to its predominance in this marketplace. That’s not the same as naming China specifically, though some vehemently argue it is a backhanded way to do so. An aim of the ban is also to encourage U.S. manufacturing of humanoid robots. The idea is that by banning foreign-produced ones, the U.S. will see more companies set up shop in America to do this. The supply chain and industrial base for American humanoid robots will be bolstered. Our reliance on foreign producers for providing humanoid robots will be lessened. This direction makes sense since U.S. consumers and businesses are forecasted to widely and deeply embrace AI humanoid robots over the next several years. It is a huge and untapped market. One important point of clarification is that even if you purchase an American-made humanoid robot, this does not guarantee that the robot won’t spy on you or that it won’t be commandeered remotely. This can still arise. You ought to be ever vigilant about allowing humanoid robots into your private spaces. The belief is that at least you will have a greater chance of legally going after the robot maker if the robot does encroach on your legal rights. You can turn to the U.S. courts. The robot maker might be violating U.S. laws in your state or at the federal level and face criminal penalties. The possibility of a civil lawsuit is also on the table. The whole kit-and-caboodle is a head-scratcher because there is still the chance of letting the horse out of the barn. Things go like this. You innocently and excitedly buy a humanoid robot, and, unbeknownst to you, it leaks all sorts of privately collected information about you to a U.S. company. That U.S. company uses this information for its gain. Or maybe they post it in a database that gets hacked. The bottom line is that no matter which way things go, you have exposure, and few are thinking about this when they opt to buy these amazingly convenient devices. A final thought for now. The famous playwright Publilius Syrus made this key remark: “He is most free from danger, who, even when safe, is on his guard.” AI humanoid robots are going to become an essential part of our daily lives. It is inevitable. Keep your wits with you and keep your guard up. That robot in the corner might be watching your every move.

JAKARTA, Indonesia, July 30, 2026 (GLOBE NEWSWIRE) -- As the AI era continues to accelerate, the cockpit - the most frequently used and highly perceived human-vehicle interaction interface - has become a critical factor in shaping brand differentiation and localized experiences in the global automotive competition. On July 27, 2026, OMODA & JAECOO officially unveiled its next-generation Super AI Cockpit in Southeast Asia at the "SUPER AI NIGHT” event in Indonesia. Equipped with ByteDance’s Seed LLM, the cockpit becomes the first intelligent cockpit system in Southeast Asia powered by this advanced large language model. Powered by three-industry-leading technologies - shaping the future of AI cockpit worldwide, OMODA Super AI Cockpit achieves three breakthrough milestones: First in ASEAN to launch ByteDance’s Seed LLM in a vehicle, First automaker to bring AI agents to every driving scenario, and First AI cockpit with an advanced Master-agent architecture. Through deep localization across five major languages and the collaborative operation of 10 intelligent agents, it represents Southeast Asia’s first intelligent cockpit solution with large-scale deployment and comprehensive native AI capabilities. It enables vehicles to evolve from tools that simply "follow commands” into intelligent companions that can truly "understand user needs.” The debut model equipped with this technology, OMODA 4, also made its impressive appearance at the event. Redefining Human-Vehicle Interaction: Three Core Technology Foundations Enable a Generational Advantage A truly AI-native cockpit is not simply an accumulation of functions, but a fundamental upgrade at the architectural level. Addressing the challenges of traditional in-vehicle voice systems - including mechanical interaction, one-way command execution, and limited contextual understanding - OMODA has developed a fully integrated and independently controlled technology foundation, with the cockpit powered by three core technology systems working together. Among them, the ByteDance Seed large language model delivers advanced natural language processing capabilities, accurately supporting complex multi-round conversations, ambiguous intent recognition, and multimodal content generation such as AI wallpapers. The cloud-edge collaborative architecture combines powerful cloud computing capabilities with low-latency edge-side responses. Even in offline or weak network environments, the edge-side engine ensures stable operation of key functions including vehicle control, navigation, and offline audio and video entertainment, effectively addressing the industry challenge where intelligent systems become unavailable once network connectivity is lost. Meanwhile, Agentic AI technology drives the collaboration of 10 intelligent agents, enabling the cockpit to evolve from passive responses to an AI super assistant capable of proactively anticipating user needs. From "Easy to Use” to "Understanding You”: Three-Stage OTA Evolution Enables Long-Term Lifecycle Growth A truly useful AI system must continue to evolve. Through a three-stage OTA evolution roadmap, OMODA enables its Super AI Cockpit to become increasingly personalized, adaptive, and dynamic over time. In the currently deployed AI 1.0 stage, 10 intelligent agents are fully integrated into the cockpit. The Orchestrator agent supports complex intent decomposition, while the Vehicle Control Agent supports up to 7 vehicle control commands within a single user utterance. Covering navigation assistance, vehicle usage guidance, conversational companionship, real-time information services, and personalized wallpapers and user name customization, AI 1.0 delivers a full-scenario experience where users can control functions effortlessly through voice interaction. The AI 2.0 stage, planned for launch in 2027, will focus on four key dimensions: Better Intent Understanding, Better Understanding of User Habits, Better Understanding of Emotions, and Better Understanding of Trends, further upgrading the cockpit from "more convenient” to "truly understanding you.” At this stage, the cockpit will be capable of simultaneously handling more than eight complex intents, allowing users to complete route planning and multi-stop navigation through a single sentence. Powered by a scenario-aware intelligent engine, the system can automatically activate dedicated modes like rainy weather mode based on real conditions. By integrating ANC noise reduction, AI Voiceprint Cloning, AI Pet Replication, and dynamic wallpapers, the cockpit will create a more emotional and personalized cabin environment. It will also deeply integrate the O4 in-vehicle gaming ecosystem (equipped with a game controller) and AI smart sound tuning capability, delivering engaging lifestyle experiences for younger drivers. In the ultimate AI 3.0 stage, the intelligent cockpit will integrate the AI Box high-performance computing chip platform and deploy an on-device multimodal large language model, forming a dual-engine architecture that delivers millisecond-level edge response alongside in-depth cloud reasoning. This will truly bring an AI brain into the vehicle, creating an intelligent assistant that accompanies users anytime and anywhere. Partnering with Tech Leaders to Secure Global Leadership in the Next Phase of AI Mobility As a pioneer in the global young drivers segment, OMODA is leveraging its self-developed full-stack underlying architecture to establish a deep strategic collaboration with BytePlus, the enterprise technology service brand under ByteDance. By combining OMODA’s advanced intelligent cockpit capabilities with BytePlus’s parent company ByteDance’s deep understanding of the digital mindset and cultural trends of global young drivers, cater to the travel habits and interaction preferences of global young drivers. Together, they will create an intelligent mobility companion featuring both emotional intelligence and personalization, pioneering a new model of collaboration between intelligent vehicles and leading AI technology companies. In the future, both parties will continue exploring innovative applications of the Seed large language model across diverse mobility scenarios. Choosing Indonesia as the location for the regional debut in Southeast Asia represents a strategic milestone for OMODA & JAECOO in capturing the core voice of intelligent transformation among young markets in Southeast Asia and around the world. Going forward, OMODA 4 will serve as an important vehicle platform for this technology, rapidly expanding into key global markets. Driven by AI innovation, OMODA & JAECOO will continue building the leading global intelligent cockpit perception and accelerate the industry’s transition toward a new era of intelligent mobility. OMODA 4 About OMODA & JAECOO As a youthful, personalized global brand, OMODA & JAECOO lives by the vision of "Co-creating a Beautiful Life with Young People.” OMODA is dedicated to embracing pioneering global consumers, striving to build "the World's Leading Crossover Brand.” It delivers fashion-forward vehicles with cutting-edge design and futuristic technology to Gen Z, redefining trendy travel culture with a crossover attitude. JAECOO adheres to the philosophy of "From Classic, Beyond Classic,” and is committed to becoming a "Global Elegant Off-Road Brand.” With exceptional four-wheel drive performance, forward-thinking intelligent technologies, and outstanding safety features, it leads a new era of elegant off-road driving. Born Green, OMODA & JAECOO leverages the world-leading SHS super hybrid technology (covering both PHEV and HEV) to offer the best hybrid solution for global users, driven by the core advantages of "Super High Power, Super Low Energy Consumption, and Super Long Combined Range.” At the same time, the brand is accelerating its BEV technology deployment, responding to diverse mobility needs with stronger and more comprehensive technological capabilities. In terms of intelligence, the brand focuses on intelligent driving and intelligent cockpit. Powered by SIVP (Super Intelligent Valet Parking) and AI cockpit technologies as key enablers, it builds a full-scenario smart mobility experience and continues to lead the future of mobility. In addition, in collaboration with AiMOGA, the brand has developed robots that extend smart technology into diverse interactive scenarios, broadening the landscape of smart living. Driven by deep insights into user needs, the brand hit one million in sales in just three years, setting the fastest growth record in the global automotive industry. To date, it has expanded into 77 markets worldwide, covering Europe, Asia, Australia, Africa, Latin America, and the Middle East. Europe stands out as a particularly strong market - the brand has already entered 22 European countries and become one of the fastest-growing car brands. Contact Person: Wu Zehui Email: [email protected] Website: https://www.omodajaecoo.com/ Photos accompanying this announcement are available at: https://www.globenewswire.com/NewsRoom/AttachmentNg/cf4c979a-420f-4ce7-b94e-b784f0fde2e4 https://www.globenewswire.com/NewsRoom/AttachmentNg/84fac60c-aa2a-432d-b03e-87a346666753 https://www.globenewswire.com/NewsRoom/AttachmentNg/7b4256a8-b691-4022-b73d-61e7865e1e79 https://www.globenewswire.com/NewsRoom/AttachmentNg/68e7cf0d-85dd-41e5-8c0b-4e239064a1ec

Posts from this topic will be added to your daily email digest and your homepage feed. See All Tech Posts from this topic will be added to your daily email digest and your homepage feed. See All Entertainment Posts from this topic will be added to your daily email digest and your homepage feed. See All News Literally ‘smashing’ — the bots can’t stop without colliding into a crash pad. Literally ‘smashing’ — the bots can’t stop without colliding into a crash pad. Posts from this author will be added to your daily email digest and your homepage feed. See All by Jess Weatherbed Posts from this author will be added to your daily email digest and your homepage feed. See All by Jess Weatherbed The 9.58-second 100-meter dash record set by Usain Bolt in 2009 has been outpaced by Chinese robots participating at the World Humanoid Robot Games in Beijing. In a preliminary heat on Saturday, Tiangong Ultra, made by the Beijing Humanoid Robot Innovation ​Center, ran the distance in 9.39 seconds, followed by the Honor-developed Lightning at 9.47 seconds. It’s one of several notable achievements to take place at the event, which can best be described as a sort of annual robot olympics that’s designed to showcase advancements in bipedal bot technology. After launching in 2025, 2,056 robots from 16 countries are participating in this year’s World Humanoid Robot Games, which includes categories for sprinting, jumping, football, boxing, and more. Tiangong Ultra also ran the 400-meter race in 38.16 seconds — beating the 43.03 second world record set by Wayde van Niekerk in 2016. Unlike human athletes, however, these sprinting bots don’t typically get to perform a victory lap after crossing the finish line. Several videos show the robots colliding into cushioned walls at impressive speeds, where they (or the pieces of what’s left of them) get collected and carried away on a stretcher. Watching the various robot models actually run is both amusing and a little unsettling, given the bizarre way they move to dash so quickly. Posts from this author will be added to your daily email digest and your homepage feed. See All by Jess Weatherbed Posts from this topic will be added to your daily email digest and your homepage feed. See All Entertainment Posts from this topic will be added to your daily email digest and your homepage feed. See All News Posts from this topic will be added to your daily email digest and your homepage feed. See All Robot Posts from this topic will be added to your daily email digest and your homepage feed. See All Sports Posts from this topic will be added to your daily email digest and your homepage feed. See All Tech A free daily digest of the news that matters most. This is the title for the native ad 8 Verge Score This is the title for the native ad The Verge is a part of PMX Global, LLC, a subsidiary of Penske Media Corporation. © 2026 VM Publishing, LLC. All rights reserved. Sign in to see your notifications or create an account to join the conversation.

Written by Cheng Zi Published on 18 Aug 2026 8 mins read The embodied intelligence sector may not need another world model startup. Hu Yucheng, a researcher born in 2001, founded LatentVerse anyway, even as he insists the company should not be grouped with the growing field of world model developers. “There are too many companies working on world models, and many of them have no distinguishing features of their own,” LatentVerse founder Yucheng Hu said. But he does not want LatentVerse grouped with any of them. Nor is a conventional world model what the company is trying to build. Hu is a doctoral student at Tsinghua University’s Institute for Interdisciplinary Information Sciences. In April, he left ByteDance Seed and co-founded LatentVerse with members from Tsinghua University, Nanyang Technological University, Peking University, Qwen, ByteDance, Xiaomi, and other organizations. The company is focused on what it calls embodied-native foundation models. According to 36Kr, LatentVerse has completed a seed funding round worth a nine-figure RMB sum. Investors include GL Ventures, Crystal Stream, Agibot, Robot Era, and Innoangel Fund. Over the past two years, VLA, or vision-language-action, has emerged as a mainstream technical approach to embodied intelligence. From RT-2 to the widely discussed Pi-0.5, the development path has been relatively straightforward: connect a vision-language model, or VLM, to an action output, train it on enough robot manipulation data, and have the model learn to act in the physical world. One limitation of VLA systems is their reliance on data with action labels. Collecting such data through teleoperated robotic arms costs more than RMB 1,000 (USD 148.1) per hour on average, requires specialized teleoperators, and can be inefficient. Another approach is the pure world model, which predicts future states through video generation and then maps those predictions into actions. Data can be cheaper to acquire because internet videos can also be incorporated into training. But this approach does not fully make use of general-purpose vision-language data, limiting the semantic understanding and emergent capabilities that VLMs can provide. A model trained only to predict changes in an image, for example, cannot necessarily understand what an instruction such as “hand the cup to the guest” means. LatentVerse has chosen a third path. Rather than building either a VLA model or today’s increasingly popular world action model, or WAM, it is seeking to unify VLMs and world models within the same architecture. Hu’s previous research has explored how world models can guide robot actions. In early 2024, he proposed PAD, a multimodal diffusion VLA approach. In November that year, he introduced VPP, a video-action model designed for closed-loop inference. The work has been described as foundational to later WAM development, with Nvidia subsequently developing DreamZero on that foundation. After joining ByteDance Seed, Hu led the development of the BagelVLA series of embodied foundation models. The related work was published before Physical Intelligence’s Pi07. His experience convinced him that embodied intelligence had reached an inflection point. In Hu’s view, reaching physical artificial general intelligence, or AGI, requires overcoming the sector’s data bottleneck. Neither a pure VLA nor a pure world model can do that alone, he believes. Instead, embodied systems need a foundation model that unifies understanding, prediction, and action within a single architecture. As data collection costs fall, the challenge is increasingly shifting toward obtaining enough usable data and building models that can learn effectively from it. LatentVerse is now training its first model, UTAM, short for unified tactile action model. Building on the team’s earlier UAM, or unified action model, UTAM is designed to output visual, language, and tactile signals simultaneously. The model contains four expert modules that together form an execution chain. A vision-language expert interprets the intent behind a task. A world model expert draws on skills previously learned from video to plan a new task sequence. An action expert then uses outputs from the first two modules to generate a coarse action representation, or latent action. This could correspond to a broad instruction such as reaching out and picking up a cup. Finally, a tactile expert operating at a high inference frequency translates that coarse action into precise end-effector control signals. It adjusts force and posture in real time when, for example, a finger touches the side of a cup. The first part of the process relies on open-loop execution, while the second uses closed-loop tactile correction. The idea resembles how a person first reaches toward an object, then fine-tunes the movement based on touch. The architecture reflects the scenarios LatentVerse is targeting. “Generalization, long-horizon execution, and dexterity are not three independent technical metrics. They are capabilities that robots must possess simultaneously if they are to be commercialized,” Hu said. “We believe embodied foundation models should have stronger general-purpose and generalization capabilities, including in settings such as the home.” LatentVerse’s technical approach involves collecting data across more dimensions and training for longer-horizon tasks. Target scenarios include robots cleaning hotel rooms, packing beverages in cramped spaces, and grasping soft objects in homes. Under LatentVerse’s framework, a robot operating in the real world needs two capabilities at once: cognitive intelligence and contact intelligence. Cognitive intelligence covers understanding instructions and planning steps, tasks that VLA and world models already attempt to handle. But cognition alone is not enough. Information from the final centimeter of physical contact can determine whether a robot completes an operation. A robot may need to determine, for example, whether the surface of a drinking glass is slippery or whether a soft object deforms when grasped. Vision alone cannot capture all of this information, so the robot must make real-time corrections using tactile feedback. These tasks require more than predicting what happens in the next frame. A robot must understand instructions, plan steps, predict physical changes, sense contact, and generate precise actions. Embodied intelligence systems have already demonstrated capabilities in some fixed scenarios, including certain precision operations in industrial settings. Achieving broader generalization through approaches such as VLA, however, remains difficult. One advantage LatentVerse sees in a unified approach is access to a broader range of training data than pure VLA systems can use. Manipulation videos, cross-embodiment robot data, and general-purpose VLA data can all be incorporated into the training framework. LatentVerse expects generalization across environments and objects to improve as the scale and diversity of that data increase. The constraint is equally clear: there is still not enough usable, high-quality data. Although embodied intelligence development has expanded data collection in recent years, much of that data was gathered for specific tasks and can be difficult to reuse when training general-purpose capabilities. Hu said achieving a major jump in capability along the same model path requires two things. “First, the scale of the data has to increase. Second, the model has to be able to use massive amounts of data, not just what we collect ourselves, but internet-scale video data and heterogeneous data across different embodiments,” he said. “The dimensions involved in data processing also become more complex. You need to describe how an entire task is broken down and use that information at different stages of training.” LatentVerse said it has spent the months since its founding building a pipeline spanning data collection, processing, and model training. This includes a proprietary data collection system and the embodied-native world model the company is now training. The company also sees potential advantages at deployment. Embodied intelligence systems can perform well in demonstrations but prove harder to deploy at scale, largely because their models are not robust enough for changing real-world conditions. Variations in lighting, human interference, and differences in how environments are arranged can increase failure rates outside controlled demonstrations. “The VLM expert can infer human intent, understand physical information, and make a judgment about what to do next. The world model expert can ingest massive amounts of data to improve robustness across environments, objects, and tasks, while turning discrete text into semantically rich visual signals,” Hu said. Data is a prerequisite for making an embodied-native model work, but LatentVerse’s approach also increases training complexity. Processing long-horizon tasks requires breaking them into multiple steps and annotating those steps sequentially, making the process more complex than conventional image or text processing. Training must also span multiple stages and tasks. Some WAM approaches rely on diffusion models, while inference latency remains an engineering challenge. “Our team has worked on pure VLA approaches and pure world models, so we know where the limitations of each lie,” founder Hu said. Another LatentVerse co-founder focuses on dexterous-hand algorithms and motion control, adding expertise on the control side. Once a data pipeline linking robot embodiments with human hand data is established, the company hopes to create a self-reinforcing data loop for embodied intelligence applications. LatentVerse has assembled a small team specializing in embodied intelligence algorithms, dexterous hands, simulation, 3D vision, and other areas of robotics. Its data strategy spans three categories. Teleoperation data, collected using virtual reality and carrying precise action labels, is used only to train the action expert. Human hand data, including egocentric data captured from a person’s point of view, is used to train the understanding and world model experts. Open-source video data is used for the same purpose. For data collection, LatentVerse is betting on an embodiment-agnostic approach, meaning data collection methods that are not tied to a specific robot form. The team is developing its own data collection glove and says it has established preliminary partnerships with several data collection facilities. “Existing data collection gloves still do not adequately meet our requirements,” Hu said. LatentVerse eventually plans to use wearable devices to enable nonprofessionals to contribute data, potentially integrating collection into everyday activities. The company estimates that this approach could cut collection costs from more than RMB 1,000 per hour to less than RMB 100 (USD 14.8) per hour. For Hu, however, model performance ultimately depends on both data quality and how that data is used. LatentVerse’s core members have been working on unified models since early 2024. Co-founder Zhang Jianke, for example, has led several embodied intelligence projects. They include HiRT, short for hierarchical robot transformer, a fast-slow system that was adopted by Figure’s Helix; UP-VLA, described as the first embodied model to unify generation and understanding; and VLM4VLA, which examined the limitations of VLMs as embodied foundation models. “The emergence of capabilities in embodied models cannot simply copy the development path of large language models,” Hu said. “While scaling data, we should also fully use the structured information contained in different types of heterogeneous data to build unified representations.” Under LatentVerse’s plan, the company will release its first 16-billion-parameter embodied foundation model within one quarter. Over the next 1–1.5 years, it plans to accumulate hundreds of thousands of hours of training data. LatentVerse also wants to incorporate more complex signal dimensions, including touch, into its models. If it can develop stronger general-purpose and generalization capabilities, it aims to address a broader range of industrial, commercial, and household scenarios. That strategy reflects Hu’s broad definition of a world model. In his view, any model that predicts the next token can be considered a world model. The meaningful distinction lies in how each approach defines its state space. “We believe ours is an embodied-native world model, a world model for robotics,” he said. Behind that definition is a view of where embodied intelligence can ultimately be used. Hu argues that one reason robots remain concentrated in narrow, vertical-specific scenarios is that their underlying intelligence is not capable enough. Only when general-purpose capabilities and generalization improve sufficiently, he believes, can robots operate reliably in unstructured environments such as homes. “The last centimeter embodied intelligence must cross is the gap between cognition and contact,” Hu said. Robots, in his view, cannot just learn to understand the world. They also need to learn to touch it, and to turn those capabilities into practical value in real-world environments. KrASIA features translated and adapted content that was originally published by 36Kr. This article was written by Deng Yongyi for 36Kr. Note: RMB figures are converted to USD at rates of RMB 6.75 = USD 1 based on estimates as of August 18, 2026, unless otherwise stated. USD conversions are presented for ease of reference and may not fully match prevailing exchange rates. Loading... Subscribe to our newsletters KrASIA A digital media company reporting on China's tech and business pulse.

ByJohn Koetsier, Senior Contributor. By 2026, we should have humanoid robots in private homes helping with laundry, vacuuming, and the dishes, at least in beta testing, says Peter Diamandis. By 2040, there could be as many as 10 billion globally in all areas of the economy, and their labor might be as cheap as $10 a day. “If you lease it like you lease a car, a $30,000 car, your price point per month is 300 bucks,” says author, futurist, investor, doctor, and engineer Peter Diamandis in a recent TechFirst podcast. “And that translates amazingly to $10 a day and 40 cents an hour. So you’ve got labor that’s waiting for whatever your wish is. You know, clean up the house, go mow the lawn, you know, please change the baby’s diapers.” Today of course most robotic manufacturers are focused on building tools for labor: warehousing, logistics, manufacturing. Just recently we’ve seen Digit by Agility Robotics get a paying gig, followed by Figure’s latest shipping model, Figure 02. But in the future, they’ll be everywhere in our economy, Diamandis says: in healthcare, manufacturing, the service industry, public and urban spaces, transport, even entertainment. This is such a transformational change that analysts don’t yet really understand how to estimate its value: Goldman Sachs says selling humanoid robots will be a $38 billion space by 2035, while Ark Invest says the resulting economic value of their labor could be as high as $24 trillion. The reason for the vast divergence: analysts’ opinions on what jobs humanoid robots will take. If the robots get really good, the high end of their value is Everest-like in elevation: “50% of Global Domestic Product (GDP) is paying humans to do work every day, in other words human labor,” Diamandis’ recent autonomous robot report quotes Brett Adcock, CEO of Figure AI as saying. “That amounts to a marketplace of $40 trillion a year. It's ten times bigger than all of transportation combined.” Diamandis has identified 16 market leaders and up-and-comers in the space. Here they are, along with the name of their autonomous robot: Where are these companies? Almost exclusive in the U.S. and China: six are in the United States, eight are in China, one is in the UK, and one is in Canada. None, at the moment are in the European Union, or South America, or Africa. The big question right now is: which companies are going to win the battle to provide these billions of robots? And: which countries are going to win? These two questions are inextricably linked, because winning the race to develop autonomous robots is perhaps the economic, financial, and societal battle that will decide the future, on multiple levels. On the economic and financial levels, the companies and countries that crack effective and efficient humanoid robots first will have a huge advantage in both labor force costs and labor force size: a massive deal for global economic power, especially for nations with a generally aging population. On the societal level, nations or geopolitical groupings that solve autonomous humanoid robots will also have the opportunity to remake their communities in a world in which labor costs approach zero: tricky, difficult, guaranteed to be controversial, but a puzzle that contains the seeds of unleashing human potential unbounded by the need to made widgets and move things. There’s even a military level to this: any observers of the Russia-Ukraine war know that drones, autonomous and semi-autonomous robots, and AI are increasingly the lion’s share of the weaponry that is winning on the battlefield. This is already happening: Anduril recently announced a billion-dollar investment into a hyper-scale factory in Ohio to “redefine the scale and speed that autonomous systems and weapons can be produced for the United States and its allies and partners.” The scope of the potential transformation here can almost not be overestimated. A big question is this: How much are the robots going to cost? The answer will drive who can afford to employ them: which countries, which companies (for which jobs), and which people. Diamandis thinks an equivalent of around $30,000 is where we’ll get to within a decade or so, which translates to around $10/day in leasing costs. That changes a lot: “What made China successful over the last 40 odd year is their low labor rate,” Diamandis says. “They had a lot of humans at very low cost that could manufacture almost anything ... [but] the cost of living has been going up in China, so the labor rate per hour is going up.” But not just in China. In the United States as well: “California minimum wage is 20 bucks an hour,” he adds. “How do you ever not put a robot in that spot at 40 cents an hour? Which works 24/7, no drug testing, no fights with her girlfriend or boyfriend, no sick days. I mean, it gets pretty compelling.” Another need is elder care. The UN predicts that by 2030—just five years away—the United States will have 25 people over 70 years old for each 100 people age 24 to 69 ... a dependency ratio of 25%, the report says. We desperately need safe and effective humanoid robots to help here. I’ve just personally learned how taxing it is on individuals (and by extension, society) to take care of only one aging parent. Offloading some of that labor onto humanoid robots—but ideally not the human connection that the elderly still need—would be a huge help. Perhaps the biggest question is what kind of world we want to build in a post-labor society. Ideally it’s one summarized by this quote from the Indian thinker Sadguru: “Technology is the means by which humanity takes a vacation from basic survival.” But there are plenty of other ways autonomous robots could go too, and we see both of them in the Russia-Ukraine war as well as the deepening financial divide in America.

Nick Huber PublishedMarch 26 2024 Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter. From steam engines to conveyor-belt assembly lines and robots on the factory floor, the manufacturing industry has long been a pioneer of new technologies. Artificial intelligence now looks set to become the next and, perhaps, biggest leap forward. But what will it mean for jobs over the next decade? Manufacturing is already highly automated, with sensors, software and computer networks monitoring the output, data, pressure and temperature of factory machines and industrial processes. Such connectivity has become essential on sites that are sometimes square miles wide. “In a refinery or petrochemical plant, there can be thousands — if not tens of thousands — of instruments, equipment and valves [needed] to, for instance, manage 250,000 to 500,000 barrels of oil per day and process that into gasoline,” points out Jason Urso, chief technology officer in the software division of Honeywell, a US industrial conglomerate. Within 10 years, more than 80 per cent of manufacturing facilities could be using AI to help run these “control systems” and fix problems with them, he forecasts. If, for example, a machine emits an unusual sound, a factory worker can ask the AI software to analyse that sound, summarise the problems associated with it and recommend remedial action, Urso says. Some manufacturers already invest in this type of AI. United States Steel Corporation, for example, has said it will use generative AI software from Google to guide its workers through truck repairs and ordering parts. AI is also playing a bigger role in product design. For example, AI-powered software can help automotive engineers make multiple three-dimensional vehicle designs in minutes instead of days, says Stephen Hooper, vice-president of software development, design and manufacturing at US software supplier Autodesk. “You [can] build 3D [computer designs] of styling for new vehicles in a fraction of the [current] time,” he notes. “You can control characteristics like the wheelbase, the vehicle type . . . and [the AI] will derive hundreds, if not thousands, of alternatives”. Hyundai has used Autodesk software to help design parts for a prototype car, the wheels of which can transform into four legs to walk and climb — making it a potential rescue vehicle. In factories, robots have long been used to assemble parts but the next generation will be AI-powered “humanoid” robots, working alongside humans. These will have sufficient dexterity and learning capabilities to handle processes such as picking and sorting items into categories, experts say. Early versions could be operating within the next five years, predicts Geordie Rose, co-founder and chief executive of Canadian start-up, Sanctuary AI, which is aiming to create the world’s first robots with “humanlike intelligence”. Its latest Phoenix model is 5ft 7in tall, weighs 70kg, and can walk at up to 5km/h. It is operated by humans but, Rose forecasts, will eventually mimic human memory, sight, sound and touch. Demand for such humanoid manufacturing robots is going to be “significant”, according to a recent Goldman Sachs research note — particularly in electric vehicle assembly. “The central premise of this approach is that you can build a machine that’s humanlike in the way that it understands the world and acts on it,” explains Rose. However, building a machine that can react like a human “is obviously a lot harder than [building] a machine that . . . can do a couple of things that people can do”. Sanctuary’s robot can already sort mechanical parts as fast as a human, but even Rose acknowledges further improvement is needed. “The question is, how long it will take [for our robots] to go from the lab to being on the factory floor,” he says. “And that’s a very difficult question to answer.” Eventually, robots equipped with artificial general intelligence (AGI) — the same level of intelligence as a human — will be capable of designing and making things, Rose predicts. “You could ask a sufficiently powerful AGI [robot] to design a new battery and then manufacture it.” Adding AI to manufacturing robots — which do not demand pay rises or go on strike — has the potential to make millions of traditional manufacturing roles redundant. Pascual Restrepo, associate professor at Boston University and an expert on industrial robots, points out that non-AI robots have already replaced between 6mn and 9mn manufacturing jobs, globally, since the 1980s. About 500,000 of these were in the US alone. Now, most experts predict AI will lead to more jobs cuts in manufacturing. When technology leaders around the world were surveyed last year by recruitment company Nash Squared, they estimated that 14 per cent of jobs in manufacturing and automotive industries will be lost due to “automation” technologies, including AI, over the following five years. Production-line workers, quality-control assessors and machine operators seem most at risk of being replaced by AI. Gabriele Eder, head of manufacturing, industrial and automotive at Google Cloud, Germany, suggests that, in these areas, AI-powered machines and equipment can “often operate with greater precision and consistency than human operators” — requiring less human intervention in manufacturing processes. “Our members are very much worrying [about AI taking their jobs],” says Kan Matsuzaki, assistant general secretary at IndustriALL, an international union representing more than 50mn workers in mining, energy and manufacturing. He adds, however, that his members acknowledge the possibility of AI bringing benefits, such as improving safety in manufacturing. Training manufacturing workers in applying alongside AI could help them adapt and minimise job losses, but opportunities may be limited. “When you reach like 55 years old . . . can those workers be retrained to become [an] AI machine . . . specialist, for example?” says Matsuzaki “[It] is very difficult to do that.” However, some experts predict that AI will create more new jobs in manufacturing than it eliminates. They note that manufacturing companies are keen to hire, rather than fire, workers — but are hampered by a global shortage of people with manufacturing skills. New AI-related manufacturing jobs will include running AI machines, monitoring their performance, programming robots, and working in “cross disciplinary teams” with equal expertise in data science and manufacturing, experts predict. At the same time, old jobs will change and become more tech-focused, rather than being replaced by AI, says Marie El Hoyek, an expert in AI and industrials at consultants McKinsey. “Some of the manufacturing roles will need to evolve,” she says. “I imagine [in the future] you would need digital champions who are core manufacturing people but know how to translate their needs and their work into digital language to the digital team and say ‘this is what I need you to resolve’.” AI will increase demand for “forensic AI scientists”, typically from a technology background, who analyse the performance of AI systems, says Cedrik Neike, chief executive of digital industries at German technology company Siemens. “[We] need to have experts which [understand] where things go wrong to fine tune them,” he says. How widely these AI systems are deployed remains open for negotiation, though. “The ultimate question is, who will benefit from this AI?” says Matsuzaki. “If you introduce AI and automation robot[s] in manufacturing workplaces . . . you can reduce your number of workers, which means the productivity will gain [and] profit will gain . . . But there’s nothing for the workers.”

DevOps.com August 18, 2026 by Scott Robohn Summary: Scott Robohn explores how UiPath Test Cloud mitigates the “release gap” caused by rapid, AI-driven development by offering a platform that blends a combination of deterministic automation, AI agents, and human oversight. By distinguishing between deterministic “robots” for efficient, cost-effective regression and flexible “agents” for complex reasoning, the platform empowers quality engineers to use agentic reasoning selectively while managing AI-related consumption and cost. Drawing on his network engineering background, Scott advocates for adopting these software testing principles to build greater resilience and adaptability within NetDevOps and broader IT infrastructure domains. I had the pleasure of serving as a delegate for the UiPath Test Cloud Tech Field Day Showcase. I’ve heard a lot about UiPath from my consulting clients and in the market over the last few years, so I jumped at the opportunity to be a part of the delegate panel and hear directly from UiPath. Now, let me be clear about something important up front: UiPath is not presenting itself as a NetOps tool for network device config testing or network automation script testing. I say this because you may know me from my consulting work in networking, network automation, and conferences and events that all focus on network infrastructure. Over the last few years, coming from a “trad” network engineer and architect perspective, I’ve reached the conclusion that network operations and engineering is downstream from software engineering and deployment. DevOps principles make a lot of sense when applied appropriately to Networking, hence the term NetDevOps; leveraging tools like GitHub and other SCCS tools, processes and frameworks like CI/CD, and other tooling are all making lots of sense for NetOps. As a logical follow-on to that, I’m going out of my way to learn how software testing tools like the UiPath tool suite and principles from the software world can be applied to testing and managing network device config and scripts. UiPath and the delegate panel really helped shed some light on all this for me. With agentic AI becoming an increasing force in code generation, there is a real and growing “Release Gap” imposed by this ability to write more code quickly. You may have heard people say “Code is Cheap”; but keep in mind that thinking and architecture are NOT cheap. As a result, enterprises struggle to release software with confidence due to slow regression cycles (often 4-6+ weeks) and are still heavily dependent on manual testing. Being able to ship code faster with confidence is clearly a challenge and need. Josh Duke of UiPath and I had some good interaction over this issue during the showcase. Josh confirmed that the “release gap” is growing because increasing frequency of AI-generated app changes daily or weekly, putting more pressure on quality assurance teams that are, in most organizations, not fully integrated into the modern development lifecycle. He also confirmed that this pressure makes maintaining release confidence more difficult. Determinism vs non-determinism is an important issue in both software testing and NetOps. Why tolerate non-deterministic mechanisms when I need a deterministic outcome? And why burn tokens on agents leveraging an LLM when you can have a deterministic script/robot run a reliably repeatable procedure? We’re learning to build systems that allow the user to leverage both reasoning (non-deterministic) and reliability (deterministic) to get to the best outcomes in the best ways. And platforms that allow me to control what approach I use in different parts of a workflow are critical. We had a good discussion about UiPath’s capabilities around “robots” and “agents.” Josh described robots as deterministic automations, similar to the Python scripts we use in NetOps, that follow predefined logic and produce consistent outcomes without requiring LLM inference during the deterministic execution itself. I definitely prefer deterministic results for specific functions. We’re also finding that agentic reasoning adds value where flexibility is required, much like a Roomba learning to navigate around a couch or coffee table that has just been moved. Josh noted that UiPath Test Cloud gives users the flexibility to choose deterministic automation, agentic execution, or a combination of both, depending on the scenario. Even though the UiPath tool suite is focused on software dev and test use cases, I got a lot out of the discussion from a strategy perspective. It got me thinking about more connections between software development and tooling and processes for NetOps, reinforcing the connections I see between the disciplines. This is one of many aspects I love about Tech Field Day events across different IT domains: this cross-fertilization we get from different tooling, approaches, and emergent design patterns. I also see greater opportunity – and need – for us to become systems thinkers regarding our IT stacks versus just separate tech silos. Being able to span different tech categories to better understand how things connect and see application of principles across different categories really helps me think more holistically about IT infrastructure. And I have a growing hunch that AI – as both a thought/architecture partner and more powerful operations tools – will allow us to elevate our thinking and see the whole system more holistically and clearly. From my background in network architecture and automation, the hybrid approach between deterministic robots and agentic reasoning is one of the most important aspects of UiPath Test Cloud. There is significant pressure in many IT domains to automate with agents and LLMs, but we shouldn’t accept unnecessary variability or burn tokens on reasoning when a deterministic script or robot can execute a predefined procedure with greater predictability and repeatability I value when any solution provider gives the user the control to decide where I apply intelligence. Using agents for complex tasks with some ambiguity, while relying on deterministic robots for repeatable reliable heavy lifting, is the kind of systems thinking that is missing from many automation strategies (in networking and other tech domains). This borrowing of software testing principles into other operational domains is how we build more resilient and responsive systems. If you’re interested in hearing directly from customers and seeing these tools in action, UiPath is hosting their annual event FUSION September 22–25, 2026, at the Wynn Las Vegas. Filed Under: Social - Facebook, Social - LinkedIn, Social - X, Sponsored Content Tagged With: Agentic Reasoning, ai, Deterministic Robots, devops, NetOps, software test automation, SoftwareTesting, Solutional, UiPath August 25, 2026 | Kirubanandan R August 25, 2026 | Aiswarya Giridharan August 21, 2026 | Priya Doty August 20, 2026 | Jay Aigner August 19, 2026 | Kirubanandan R © 2026 ·Techstrong Group, Inc.All rights reserved. Modern Software Development and Delivery 1Q12Q23Q34Q45Q56Q6 How would you best describe your organization's current software delivery environment(s) on mainframe computers? (Select all that apply)(Required) Primarily manual or legacy delivery processes Automated CI/CD in limited areas Hybrid environment with multiple delivery platforms and toolchains Mostly standardized CI/CD platform Intelligent, governed software delivery platform My organization does not utilize mainframes Not sure / don't know Next Which outcomes are most important to your organization's software delivery strategy today? (Select up to three)(Required) Accelerating software delivery Automating build, test, and deployment workflows Integrating security and compliance into delivery Providing governance and auditability Supporting AI-assisted software delivery Improving software supply chain visibility Standardizing delivery across teams Supporting both modern and legacy application delivery Other Not sure / don't know Previous Next What, if anything, is limiting your organization's mainframe software delivery progress? (Select up to three)(Required) Legacy applications or legacy development processes Fragmented tools and disconnected workflows Security, governance, or compliance requirements Limited automation across the delivery lifecycle Difficulty integrating AI into software delivery workflows Skills or resource constraints Organizational or cultural resistance to change Limited visibility across the software delivery lifecycle Budget or investment constraints Other Nothing is significantly limiting our progress Not sure / don't know Previous Next In which areas of your mainframe software delivery environment are you currently using AI? (Select all that apply)(Required) Code generation / developer assistance Automated testing / test generation Build and pipeline optimization Deployment automation Monitoring and incident response Security / compliance automation Other We are not currently using AI in our environment today Previous Next As software delivery responsibilities expand beyond traditional build, test, and deploy, which of the following areas is the most challenging for your organization today with respect to mainframe software delivery? (Select one)(Required) Integrating security throughout the software delivery lifecycle Meeting governance, audit, or compliance requirements Managing software supply chain integrity and visibility Operationalizing AI across software delivery workflows Coordinating software delivery across diverse tools, teams, and environments Other None of these are significant challenges Not sure / don't know Previous Next Which of the following do you expect is most likely to accelerate your organization's mainframe software delivery progress over the next 12-18 months? (Select one)(Required) Greater automation across the software delivery lifecycle Broader use of AI-assisted software delivery workflows Stronger governance, policy, and compliance automation Improved software supply chain visibility Better integration across software delivery tools and platforms Greater standardization across development teams Other No significant changes are planned Not sure / don't know Submit Submit Δ Δ

Всем привет! Меня зовут Алексей Малафеев, я Senior SEO‑специалист, ex‑Детский мир. Также работал в качестве лида в компании KazanExpress (нынешний Магнит Маркет). В этой статье хочу показать несколько прикладных кейсов того, как можно упростить и автоматизировать рабочие процессы в SEO крупного e‑commerce проекта: от мониторинга технического состояния сайта и анализа трафика до работы с большими объёмами данных и сезонностью категорий. По опыту могу сказать, что на разных проектах встречаются как похожие и повторяющиеся задачи, так и абсолютно специфические, которые возникают из‑за особенностей конкретного бизнеса, архитектуры сайта или внутренних процессов. Но чем крупнее проект, тем чаще возникает одна и та же проблема — данных становится слишком много для ручной работы. Крупный e‑commerce — это сотни тысяч и миллионы страниц, большое количество запросов, URL, технических параметров и регулярных изменений. В какой‑то момент стандартных инструментов просто перестаёт хватать. В моём случае отправной точкой стал довольно простой момент: необходимая мне выгрузка всех проиндексированных страниц сайта перестала помещаться на одном листе Excel. А это, напомню, 1 048 576 строк. Поначалу для решения подобных задач мне хватало Power Query. Но со временем задачи становились сложнее: требовалось обрабатывать всё большие объёмы данных, сохранять историю, регулярно запускать проверки, сопоставлять между собой разные источники и автоматически получать результаты. Так я постепенно пришёл к Python, базам данных SQL, API, построению дашбордов в Yandex DataLens и Power BI. А сегодня к этому набору инструментов добавляются и ИИ‑инструменты, которые позволяют ещё быстрее решать отдельные задачи. В статье покажу несколько примеров того, что из этого получилось. Одним из первых Python‑скриптов, который я написал, был скрипт для сортировки списка URL по типам страниц на основе структуры URL. В чем была необходимость? Чем крупнее интернет‑ресурс, тем больше и разнообразнее типов страниц он может содержать. Вероятнее всего, ресурс не будет ограничиваться лишь листингами и товарами. Листинги могут делиться на страницы категорий, фильтров, брендов и так далее. Важно уметь быстро сортировать такие страницы и понимать, с каким типом страниц мы имеем дело. К примеру: site.ru/example_{pagetype_1}/ site.ru/product/{pagetype_2} site.ru/category/{pagetype_3} site.ru/category/example_{pagetype_4}/ Если проблема носит массовый характер, важно в первую очередь определить, с каким типом страниц она связана. Это позволяет при минимуме усилий найти и погасить источник «очага». Объясню на примере. Разбираем технические проблемы, которые подсвечены в Google Search Console в разделе Pages. Как мы знаем, из интерфейса нельзя выгрузить полный список страниц — есть ограничение на количество строк в выгрузке. Но даже несколько сотен или тысяч URL могут дать нам понимание того, какой тип страниц содержит наибольшее количество невынужденных технических ошибок: редиректы, 404, проблемы с canonical и так далее Скрипт показывает: типы страниц; количество страниц; процентное соотношение от общего числа страниц, которые мы отдали на вход. На основе полученных данных уже можно расставлять приоритеты: в первую очередь разбираться с тем типом страниц, который формирует наибольшую долю проблем. Да, со списками в несколько сотен или тысяч строк сейчас вполне легко справятся ИИ‑инструменты. Достаточно достаточно точно описать паттерн и дать исходные данные. Более того, современные ИИ‑агенты уже способны самостоятельно написать подобный скрипт или обработать готовый файл. Но здесь есть важный нюанс. Когда речь идет о регулярной работе с миллионами URL, автоматизация нужна уже не для того, чтобы один раз обработать файл. Нам нужно воспроизводимо выполнять одну и ту же операцию, сохранять результаты, сравнивать их между собой и встраивать этот процесс в общую систему мониторинга. И в этом случае небольшой Python‑скрипт до сих пор остается гораздо более удобным инструментом, чем ручная работа с ИИ. В стандартных отчётах Яндекс Метрики не всегда удобно получить единую картину динамики трафика по всем необходимым типам страниц. Почему стоит обратить на это внимание? Потому что релиз одной конкретной задачи может сказаться на разных типах страниц одновременно и с диаметрально разными результатами. Один тип страниц может демонстрировать прирост трафика в то время, как другой тип страниц будет показывать просадку. Чтобы иметь полное представление об изменениях, приходится применять дополнительный инструментарий. Поэтому для более детального анализа можно вынести данные за пределы интерфейса Метрики: через API получить статистику, классифицировать URL по типам страниц с помощью регулярных выражений и сохранить результат в собственной структуре данных. После этого можно построить дашборд в DataLens, Power BI или любом другом удобном BI‑инструменте. На этом этапе уже появляется возможность смотреть не только на общую динамику органического трафика, но и разбирать её на составляющие. Например, увидеть, что после определённого релиза общий трафик практически не изменился, хотя внутри него произошли заметные изменения: один тип страниц вырос, другой просел. Если вы следите за хронологией релизов (а на крупных проектах без этого никак не обойтись), то вскоре у вас будут отправные данные для анализа того, как те или иные изменения отразились на трафике разных типов страниц. Иногда резкие колебания происходят именно там, где их совсем не ожидаешь. При этом сам дашборд — скорее инструмент для поиска точки, на которую стоит обратить внимание. Дальше уже необходимо разбираться, что именно произошло: был ли это результат конкретного релиза, изменение поискового спроса, сезонность, действия конкурентов или какие‑то технические проблемы. В крупном проекте важно мониторить не только наиболее важные и трафиковые страницы, но и в целом иметь возможность отслеживать состояние большого количества страниц. А также фиксировать тот момент, когда что‑то вдруг изменилось. Закончился ассортимент не на самой популярной странице фильтра — легко. Ушёл бренд (продавец) с вашего маркетплейса — запросто. Страница в какой‑то момент в meta robots вдруг стала отдавать noindex — и такое проходили. Причин может быть уйма. Вовремя распознать проблему и предотвратить в дальнейшем подобное поведение — бесценно. Именно поэтому особое внимание я уделяю тому, как функционирует и что входит в sitemap.xml. Карту сайта я использую как один из источников информации о том, какие страницы на сегодняшний день присутствуют на сайте и, по задумке проекта, должны быть доступны для сканирования и потенциальной индексации. Внезапное исчезновение большого количества страниц из sitemap.xml без очевидных на то причин — довольно тревожный звонок. Поэтому я использую скрипт, который ежедневно парсит sitemap.xml и записывает полученные данные в базу. Каждая таблица соответствует отдельному типу страниц. URL выступает в качестве primary key, а помимо него мы сохраняем дату первого и последнего появления страницы в sitemap.xml. Таким образом, со временем формируется история изменений. Можно определить, в какой момент конкретные страницы или целые группы страниц перестали попадать в sitemap.xml, а затем уже сопоставить это изменение с другими данными и найти причину. Например, если одновременно исчезли тысячи товарных страниц, в тот же день изменился шаблон генерации sitemap и после этого в Google Search Console начала расти доля исключенных страниц, это уже совсем другая ситуация, чем исчезновение нескольких URL, на которых действительно закончился товар. На крупных проектах, конечно же, есть свой отдел тестирования и свой пул автотестов, составленных на основе технических заданий SEO‑отдела. Но по мере роста проекта увеличивается количество объектов и частота необходимых проверок. Поэтому порой проще своими силами внедрить небольшой набор специализированных проверок на Python, чем каждый раз ставить отдельную задачу на разработку и встраивание новых тестов в существующую систему. На сайте, где сотни тысяч или даже миллионы страниц, нет необходимости проверять каждую. Это будет слишком ресурсозатратный процесс и далеко не всегда даст дополнительную пользу. Достаточно выделить несколько репрезентативных страниц из каждого типа и впоследствии регулярно проверять именно их. Заголовки, ссылки, наличие текста, плитка товаров — слететь может что угодно. Поэтому важно отслеживать наличие на странице ключевых элементов, которые влияют на on‑page SEO и перелинковку со стороны фронтенда. При этом набор тестовых страниц не обязательно должен быть статичным. Если структура сайта меняется, появляются новые типы страниц или меняется шаблон существующих, набор URL для проверок тоже необходимо пересматривать. Иначе со временем можно получить автотесты, которые исправно проверяют уже не самую актуальную версию сайта. Для сайтов на SSR также немаловажно тестировать страницы с разными User‑Agent. Может быть такое, что поисковые роботы получают либо слишком урезанную версию страницы, либо, наоборот, слишком много ненужного контента. Поэтому полезно контролировать не только то, что видит обычный пользователь, но и то, какую версию страницы получают Googlebot и YandexBot. Для удобства мониторинга я использую уведомления в мессенджер, куда на постоянной основе приходят сообщения о состоянии тех или иных страниц и параметров. В результате получается довольно простая схема: выбираем репрезентативные URL → регулярно проверяем ключевые элементы → сравниваем результаты с предыдущими запусками → при отклонениях отправляем уведомление. Запуск скрипта можно настроить как локально, так и на сервере с помощью cron. В любом проекте SEO‑задачи должны помогать бизнесу добиваться своих целей. Поэтому помимо своих непосредственных обязанностей полезно контактировать и делиться данными со смежными отделами. На маркетплейсе разные категории имеют разную специфику и, соответственно, разную сезонность. Для одних товаров пик спроса приходится на Новый год, для других — на начало учебного года, дачный сезон или определённые погодные условия. Если коммуникация с коммерческим отделом налажена, такая информация может быть полезна не только SEO‑команде. Например, понимание того, когда начинает расти спрос на конкретную категорию, может помочь заранее подготовить ассортимент, запланировать закупки и обратить внимание на наличие товаров. Для этого я использую следующий подход: В рамках каждой вертикали собираем сопоставимый шаблон семантики с коммерческими запросами вроде «купить», «цена» и так далее Собираем исторические данные по частотности запросов за максимально доступный период. Нормализуем данные и сравниваем динамику между категориями. Строим дашборд с учётом структуры каталога, чтобы сезонность было удобно анализировать не только по отдельным запросам, но и на уровне категорий и вертикалей. Регулярно обновляем данные, чтобы постепенно накапливать собственную историю и видеть изменения спроса из года в год. В этой статье я привёл несколько примеров того, как с помощью небольших инструментов можно решать вполне конкретные задачи технического SEO: классифицировать URL, отслеживать изменения в sitemap.xml, контролировать состояние страниц, анализировать трафик по типам страниц и работать с сезонностью категорий. При этом сами инструменты здесь не являются самоцелью. У каждого проекта свои особенности, ограничения и болевые точки. Где‑то достаточно обычного SQL‑запроса, где‑то имеет смысл написать небольшой Python‑скрипт, а где‑то уже понадобится полноценный дашборд в Power BI или Yandex DataLens. Поэтому я стараюсь смотреть на задачу немного шире: искать повторяющиеся процессы, находить точки, которые можно автоматизировать, и не бояться самостоятельно собирать инструменты под конкретную проблему. Не обязательно каждый раз изобретать сложную систему. Иногда несколько строк SQL или небольшой Python‑скрипт могут сэкономить часы ручной работы и, что гораздо важнее, позволить регулярно контролировать то, что раньше проверялось только время от времени. Именно такой подход я стараюсь применять в своей работе: сначала найти проблему, затем понять, какие данные для её решения нужны, и уже после этого выбрать подходящий инструмент. P. S. В качестве бонуса прикрепляю ссылку на GitHub, где публикую свои SEO‑скрипты: мониторинг мета‑тега robots, отслеживание изменений robots.txt, парсер sitemap.xml и другие инструменты, которые использую или дорабатываю в работе. Senior SEO | ex — SEO Team Lead Детский мир Начиная примерно с вечера 26 августа, на ТСПУ стали перехватывать DNS запросы к крупным DNS серверам CloudFlare и Google (1.1.1.1, 8.8.8.8), ранее блокировали DoH сервера от данных корпораций. Результат DNS резолвинга выглядит следующим образом: Ваш аккаунт Разделы Информация Услуги

We spoke to Luis Pacheco about the story behind Animaquina, from early experiments rigging robots with Blender's Python API to its real-world applications and how Geometry Nodes became the technical core of the project. My name is Luis Pacheco, though most people know me as Luigi. I'm an Assistant Professor of Architecture at Florida Atlantic University's School of Architecture, where I direct the Interactive Machines Lab. I'm an architect by training, but I've always been interested in technology and how things get made. I got into digital fabrication in the early 2010s when I co-founded MakerMex, one of Mexico's first desktop 3D printer companies. We built 3D printers, laser cutters, and educational tools around STEAM learning. One of those projects was MakerScad, a visual programming platform my business partner developed to make 3D design more approachable for kids. Building open tools has always been a big part of my career. My Blender story actually started in architecture school. My Windows PC crashed right before a final, and the only thing I could get running on Linux fast enough to finish my missing renders was Blender. I never looked back, and it became my home base for everything 3D. So when I started working with industrial robots, I naturally wanted to bring them into Blender. That curiosity eventually became Animaquina, which grew into the core of my doctoral research. It started pretty organically. Early on, I was rigging robots and experimenting with Blender's Python API through lots of small prototypes and little experiments. When I began my Doctor of Design, I unified all of those ideas into a single interactive robotics environment. I emphasize interactive because that's where Blender really changed the game. What surprised me along the way was how much animation has to offer robotics. Blender's IK, constraints, drivers, and Geometry Nodes turned out to be a surprisingly powerful robot programming toolkit; it just wasn't labeled that way. With Blender's animation tools, you can tell one robot to look at another, track a person, or respond to a moving object in real time. That kind of reactive behavior is genuinely difficult to achieve in traditional robot programming software. In Blender, it's almost trivial because animators have been solving these kinds of problems for decades. I didn't start out trying to reinvent robot programming. I simply wanted to replace my Rhino, Grasshopper, and KUKA|prc workflow because I was already spending most of my time inside Blender. Originally, I treated toolpaths as a static list of points that a robot followed one after another. Blender pushed me beyond that. Once your robot lives inside a real-time scene, toolpaths stop being static and become behaviors that you can animate, modify, and react to while everything is running. It's already being used in production. Animaquina has been used across multiple research labs, but it's also been used on real built projects. The Dancing Columns project with Joseph Choma, exhibited at the Phase Gallery in Miami Beach, was entirely fabricated with Animaquina. More recently, a six-meter-tall column by artist Edouard Duval-Carrié, currently on display at the Venice Biennale, was produced using it for both the robotic 3D printing and robotic milling. Today, Animaquina is in open beta, and besides the beta users, I use it daily for research across KUKA, Universal Robots, and xArm platforms. While the project started around robotic fabrication, I'm increasingly interested in motion control, entertainment, and robotic cinematography. I'm excited to see what that community builds with it. This is a light painting with the robot made by students in a course I give to master's students at Universidad de las Americas in Quito, Ecuador. Geometry Nodes has become the foundation for building custom procedural fabrication workflows. I use it to slice surfaces, rearrange points to avoid collisions, generate toolpaths, visualize the result, and simulate the robot before anything happens in the real world. The workflow itself is straightforward. You add a robot, select its brand and model from the available rigs, and connect to the physical machine. Once connected, Blender continuously reads the robot's state, so its joint positions and TCP are mirrored inside the scene. Move the real robot, and the digital twin updates instantly. That's incredibly useful when mapping real-world objects (like a 3D printing surface) into Blender. To program the robot, you simply move a target Empty. Position it, rotate it, and command the robot to follow it. You can also assign a curve, and the robot will automatically follow that path. This is where Geometry Nodes becomes really powerful. Every point along the path can carry attributes such as speed, extrusion rate, cooling fan values, or any custom parameter you define. The fabrication logic travels with the geometry itself instead of living somewhere else in the software. For robotic 3D printing, custom node groups slice the geometry using different strategies. Traditional directional slicing follows the Z axis, while non-planar and geodesic slicing follow the surface to improve overhangs. Multi-axis strategies also reorient the robot's tool continuously to optimize printing. At this point, I've built node groups for portrait drawing, vase printing, milling, non-planar printing, and several other fabrication processes. Because everything is procedural, changing the design automatically updates the robot path. Tweak a curve, and the toolpath, the simulation, and ultimately the physical robot all update with it. Then there's Puppet Mode, where Blender's animation tools really shine. Animate the target with a Follow Path constraint (or any other animation tool), and the physical robot follows in real time. Because the robot is driven directly from Blender's animation system, iteration becomes incredibly fast. The area I'm most excited about now is bringing live data into Blender. I'm developing an add-on that exposes OSC and MQTT data as Blender properties, allowing Geometry Nodes to react to live sensor information. The same workflows that have driven procedural animation for years can now drive physical robots. It's still a work in progress, but I think it opens the door to truly real-time procedural robotics. Besides Animaquina itself, I've developed FabNodes, a companion add-on for generating G-code and fabrication workflows using Geometry Nodes attributes. It's heavily inspired by Alessandro Zomparelli's G-code Exporter, which showed how powerful Blender could be for digital fabrication. I've also built an MQTT add-on that lets Blender both receive and publish MQTT messages. MQTT is a lightweight messaging protocol widely used in IoT, and it's how Blender communicates with end effectors such as motors, LEDs, sensors, cameras, and other hardware. Because it's hardware agnostic, the same workflow works whether I'm controlling a KUKA, Universal Robots, or xArm robot. One of Blender's greatest strengths is that it ships with a full Python interpreter. That makes it straightforward to integrate almost any Python library, and it's one of the reasons Blender has become such a flexible platform for building a complete robotics development environment. The biggest challenge has been supporting multiple robot manufacturers. Every brand has its own programming language, communication protocol, controller, and limitations. A huge amount of the development work goes into hiding that complexity behind a single interface, so artists and designers can focus on creating instead of learning the quirks of each robot. The most unexpected part wasn't a solution; it was a discovery. I kept running into robotics problems, reaching for whatever Blender had, and realizing it already solved them. Constraints, armatures, drivers, Geometry Nodes, and keyframes kept solving robotics problems they were never designed for. It feels like the tools were already there, waiting to be used differently. I think we're at a point where a single developer has access to tools that simply didn't exist or weren't as accessible ten years ago. Blender, Python, GitHub, AI assistants, and open-source communities have dramatically lowered the barrier to building software. Animaquina itself started as a side project during my doctoral research. My advice is not to try to build the whole thing at once. Pick one problem you're genuinely interested in solving, and keep building from there. Animaquina is currently in open beta. At the moment, it supports KUKA, Universal Robots, and UFactory robots, and I'm currently working on adding ABB support. Anyone interested in trying it can register for the beta here. I'll continue sharing updates, tutorials, and development progress as new features are added. Start receiving our weekly newsletter Facebook Twitter YouTube Instagram Podcasts © 2026. 80 level. All rights reserved

From daily news and career tips to monthly insights on AI, sustainability, software, and more—pick what matters and get it in your inbox. Access expert insights, exclusive content, and a deeper dive into engineering and innovation all with fewer ads or a completely ad-free experience. All Rights Reserved, IE Media, Inc. Follow Us On Access expert insights, exclusive content, and a deeper dive into engineering and innovation all with fewer ads or a completely ad-free experience. All Rights Reserved, IE Media, Inc. COMPASS adapts NVIDIA’s X-Mobility policy with RL, while AI agents automate testing, training, diagnosis, and evaluation. Researchers have developed an AI agent-driven approach that could make it easier to train robot navigation systems across different machines and environments. NVIDIA’s new framework, named COMPASS, is designed to reduce the time and effort needed to adapt navigation policies when robots, scenes, or operating conditions change. The system combines AI agents with simulation, reinforcement learning, and automated testing to streamline the development process. The researchers demonstrated the approach using a quadruped robot, showing how the system can support navigation training and evaluation while keeping humans involved at key decision points. COMPASS (Cross-Embodiment Mobility Policy via Residual RL and Skill Synthesis) combines a pretrained navigation model with reinforcement learning and AI agents to adapt robot behavior without building a navigation system from scratch for every robot and environment. Robot navigation is more complicated than simply making a machine move. A navigating robot must understand its surroundings, determine where it is, choose a path, avoid obstacles, and safely reach a target. When the robot or environment changes, developers often need new data, simulation environments, software interfaces, training, and testing. Repeating this process can be time-consuming and difficult to reproduce. COMPASS aims to reduce that workload. Instead of retraining a navigation system from the beginning, it starts with NVIDIA’s pretrained X-Mobility policy. It uses reinforcement learning to train a specialist for a particular robot and environment. The specialist learns to correct the existing policy so it can better handle the physical characteristics and surroundings of the target robot. The researchers also use AI coding agents to automate much of the development process. A developer can specify the robot, environment, and navigation goal, while the agent checks software dependencies, prepares simulation assets, runs initial tests, starts training, investigates failures, and compares trained models. Human approval remains part of the process, with developers deciding whether a scene is ready, whether initial tests are successful, and whether a trained model should be promoted. The reference workflow uses the Boston Dynamics Spot quadruped robot. Developers can begin with a built-in warehouse environment, making it easier to test the system before moving to more complex settings. COMPASS can also use scenes from NVIDIA’s SAGE-10K dataset, which contains 10,000 generated indoor environments covering 50 room types. For environments that need to resemble real locations closely, the workflow can use NVIDIA Omniverse NuRec to reconstruct captured spaces for simulation. This allows developers to test and fine-tune navigation policies in environments that more closely match where a robot could eventually operate. Training begins with a small smoke test to make sure the robot, environment, cameras, and control system work together correctly. Once approved, larger reinforcement-learning runs can begin. The system saves checkpoints during training so developers can compare different versions instead of automatically selecting the final model. The researchers evaluate trained policies using measures such as the rate at which robots reach their goals, how often they fall, and how long they take to complete a route. The pretrained navigation policy and newly trained versions can be tested under the same conditions to determine whether the adaptation improves performance. Once a model is approved, it can be connected to a robot’s runtime system. The policy can use camera images, odometry, and a navigation goal to generate movement commands. NVIDIA’s cuVSLAM can optionally provide visual odometry when a robot does not already have suitable position and movement information. The approach could provide a more repeatable way to adapt navigation systems across different robots and environments, while keeping human oversight for important development and safety decisions. Jijo is an automotive and business journalist based in India. Armed with a BA in History (Honors) from St. Stephen's College, Delhi University, and a PG diploma in Journalism from the Indian Institute of Mass Communication, Delhi, he has worked for news agencies, national newspapers, and automotive magazines. In his spare time, he likes to go off-roading, engage in political discourse, travel, and teach languages. Premium Follow

14.08.2026, Studentische Hilfskräfte, Praktikantenstellen, Studienarbeiten Tower cranes are underactuated systems in which rapid trolley, slewing, and hoisting motions can induce payload oscil-lations. These oscillations reduce positioning accuracy, increase settling time, and represent an important challenge for the development of autonomous and robotic crane systems. In this project, we aim to develop a physics-based simulation and reinforcement-learning framework for swing-aware control of a lab-scale tower crane. The student will first further develop the existing crane simulation model in Isaac Sim. Particular attention will be given to the realistic representation of the suspended payload, pendulum motion, sensors integeration, actuator behavior, friction, and external disturbances such as wind. The physical lab-scale crane will subsequently be used for model identification and simulation calibration, enabling the estimation of important dynamic parameters and improving the fidelity of the simulation model. Based on the developed simulation environment, an Isaac Lab RL environment will be created to train a model-free controller capable of transporting the payload towarthe d desired target. Robustness against variations in payload mass, cable length, friction, wind disturbances, and modeling uncertainties will also be investigated through domain randomiza-tion. The final objective is to transfer the trained policy from simulation to the physical lab-scale tower crane. Subject 1: Physics-Based Simulation and Model Identification • Import and complete the tower crane URDF model in Isaac Sim. • Model the crane actuation, payloaddynamics, and wind disturbances. • Integrate sensors and update the simulation based on the real crane. • Perform experiments on the lab-scale crane to identify key dynamic parameters. • Tune and validate the simulation model against experimental measurements. • Develop the simulation for further integrating the second robotis system • Subject 2: Reinforcement Learning and Experimental Validation • Develop the crane reinforcement-learning environment in Isaac Lab. • Define observations, actions, reward functions, and training scenarios for accurate and swing-minimized pay-load transport. • Train a model-free RLPlease send your CV, transcript, and, if available, information about previous experience with Isaac Sim/Lab, reinforcement learning, simulation, or robotics.transfer. Timeline & Application Details: Application Deadline: 15.09.2026 Active (in-person) participation is required. Required Skills: • Proficiency in RL, Isaac sim/Lab, C++/Python. • Basic knowledge of dynamics and control theo-ry. • Experience with real robots is a plus Please send your CV, transcript and, if available, in-formation about previous experience with isaac sim/Lab, reinforcement learning, simulation, or robot-ics. Kontakt: mohammadreza.kolani@tum.de https://www.cee.ed.tum.de/ccbe/labs/robotic-fabrication-lab/ Aktuelles Neue Methode verbessert die Überwachung von Wasserständen in Küstengebieten Warum Äpfel anders auf den Klimawandel reagieren als Weizen TUM startet Programm für erfahrene Führungskräfte TUM entwickelt Einzelphotonenquellen für Quantenkommunikation TUM ist beste Universität in Deutschland Von der Sonnenfinsternis zur eigenen Forschungsfrage no events today.

During Xiaomi Group’s earnings call on the evening of August 18, President Lu Weibing confirmed that the Xiaomi Robot will make its public debut at the 2026 World Robot Conference, which runs from August 19 to 23 in Beijing. Lu was pretty clear about the priority: the most important application right now is on production lines in smart manufacturing factories, where robots can take over some of the manual work. On the technical side, Xiaomi is aiming for universality, both in the robot’s body size and in the models that power it. The company has gone with a full-size humanoid form that stands about 1.7 meters tall. Lu also made it clear that Xiaomi won’t develop the robot in isolation. Instead, it’ll tie into the broader “people, cars, and home” ecosystem, looking at ways it can work together in smart factories, intelligent transportation, and even household settings. This public reveal builds on some earlier internal testing. Back in July, Xiaomi founder and CEO Lei Jun shared results from the robots’ four-month “internship” at an automobile factory. At a self-tapping nut installation station that required dual-sided work, the success rate climbed from 90.2% to 98%. The robots later moved into new roles in the final assembly workshop, hitting 90% success rates on tasks like folding center-console side covers and handling material boxes, marking the first time they’d handled long-duration work with flexible parts. These factory trials show Xiaomi’s practical approach: get the performance solid in controlled industrial settings first, then show it more widely. When the World Robot Conference opens, this will be the first chance for the public to see how the company’s humanoid platform actually performs outside internal testing, and how it might fit into manufacturing and everyday scenarios. Don’t miss a thing! Join our Telegram community for instant updates and grab our free daily newsletter for the best tech stories! For more daily updates, please visit our News Section. (Source)

Loading article... A humanoid robot about the size of a primary school student had something to share in Hong Kong – it sang songs and spoke to people in Mandarin and English, answering whatever questions they posed and delighting the audience around it. More than 100 robots were showcased at two exhibitions starting on Monday at the Hong Kong Convention and Exhibition Center. The X2 Ultra robot from China’s prominent humanoid robot manufacturer AGIBOT Innovation (Shanghai) Technology Co was among them. When asked about its hobbies, the robot’s list went from doing sports and dancing to studying technology and listening to music. Describing the people in front of it is no challenge either: “a woman holding a phone, a woman holding a bag and a phone, a man holding a camera,” it said at one point. Calvin Chiu, the chief operating officer of Novautek Autonomous Driving, AGIBOT’s agent in Hong Kong, said that the robot can provide emotional satisfaction to humans through conversations and serve as a teacher to older adults and children. Different robots can be programmed with different personalities, too. “It would be like a friend,” Chiu said. In China, technology has evolved into an area of competition with the US, with national security implications. Beijing’s latest five-year plan vows to “target the frontiers of science and technology”. Speeding up the development of products like humanoid robots and their applications is part of the 2026-2030 plan for the world’s second-largest economy. Official data showed China had more than 140 humanoid-robot manufacturers and more than 330 models in 2025. London-based technology research and advisory group Omdia recently ranked three of them – AGIBOT, Unitree Robotics and UBTech Robotics Corp – as the only first-tier vendors in its global assessment in terms of shipment numbers. They all shipped more than 1,000 units of general-purpose embodied intelligent robots last year, with the first two companies shipping more than 5,000 units, the report said. In February, humanoid robots were among the highlights of the CCTV Spring Festival gala in China, a television show celebrating the Lunar New Year. A martial arts performance by children and robots stole the spotlight. Some Chinese exhibitors flexed their advances at the Hong Kong Convention and Exhibition Center on Monday, showing robotic capabilities that ranged from talking to humans, punching and sand painting to doing backflips and catching suspects with nets during security patrol demonstrations. Robert Chan, global strategy officer at EngineAI, based in Shenzhen, brought its PM01 robot to showcase its mobility, including doing a front flip. His company plans to launch two factories in China for mass production this year. He said that China enjoys advantages in certain areas, such as low-cost engineering. He also pointed to the pattern of sharing know-how between companies, unlike in the United States and Europe, where companies typically shield their own technology. Chan foresaw that the next stage of robotics would move towards robots featuring bodies looking like people, with more emotional exchanges and facial expressions, or even looking like they can breathe. That is about plugging the gap in robots’ interactions with humans, he said. “The warmth and emotion exchange with the human being. Besides helping humans to make the decision and helping humans to complete their task,” he said. One company in the exhibition appears to be moving towards that direction. From a distance, three women appear to be greeting guests at an exhibition booth at one corner. Up close, they turn out to be humanoid robots that could be the future of customer service and museum tour guides. Wang Zuhua, business director at Shenzhen DX Intech Technology Co, said that the company sold more than 400 robots designed with female features and soft synthetic faces. Some are already working in museums and government venues on the mainland, where they can lead guests to washrooms and offices or provide venue tours, he said. Malaysian visitor Russel Lupang was amazed by their appearances and movements. “It’s beautiful, but not real feeling,” he said. – AP Please enter a valid email address. Please enter a valid email address.

LG Electronics' Decades of Manufacturing Know-How Meets NVIDIA's Robotics Stack At LG's New Data Factory in Seoul News Summary LG Electronics Accelerates Robotics Collaboration With NVIDIA Officials from both companies also reviewed the current status of LG's Data Factory, which is scheduled to be fully operational by the end of the year. LG has deployed its self-developed, field-proven LG CLOiD™ home robots at scale to generate, collect and learn from data. The LG's Data Factory features training spaces where robots practice a range of tasks, as well as dedicated areas for validating and refining the resulting data sets. There is a replicated home environment where LG CLOiD learns and repeats cleaning tasks, and a simulated manufacturing space - modeled on LG's washing machine plant in Tennessee - where LG CLOiD units move, stack and assemble a variety of parts. The robots are also being deployed in LG CNS's logistics automation solutions and in a space where LG Innotek trains robotic hands. Data collected across these environments will be linked with NVIDIA's robotics stack, where it is augmented and synthesized into high-quality data for robot learning. For LG, the core competitiveness of physical AI is its ability to help build a "data flywheel," where high-quality data is repeatedly secured and then used to train and continuously improve robots. Integrating detailed data gathered from LG's manufacturing and logistics operations over decades with NVIDIA's robotics stack is expected to produce a world-class data flywheel that will help set LG apart from competitors. The Data Factory serves as a space where robots collect data through physical learning and as a forward base for expanding, synthesizing and amplifying robot-learning data collected from LG's manufacturing and logistics sites, and home appliances worldwide. NVIDIA's physical AI technologies - including NVIDIA Omniverse libraries, NVIDIA Cosmos open world models and the open NVIDIA Isaac robotics development platform - are used from data to deployment. Spanning four floors - one basement level and three above ground - with a total floor area of 10,000 square meters, the Yangjae facility will house several hundred robots by the end of the year. LG also plans to continue upgrading its data-learning systems. By the end of this year, training data collected directly at the facility and data synthetically generated and augmented using NVIDIA Cosmos open world models are expected to total 100,000 hours, equivalent to roughly 12 years of data. LG plans to use this data to advance its Robot Foundation Model (RFM), which underpins the performance of humanoid robots, as it continues to build its robotics competitiveness. LG has designated this year as the starting point for its robotics business push and is moving quickly to strengthen its capabilities. Last month, the company established a Robotics Business Center. Reporting directly to the CEO, the center oversees LG's companywide robotics business and has been tasked with improving execution and efficiency. The company has established a strong presence in the industrial and commercial robotics markets and plans to expand into home robotics. Backed by proven expertise in manufacturing and developing core robotic components, including actuators, LG delivers differentiated competitiveness spanning from components to finished products. Combined with its Data Factory, which collects and manages the large-scale data needed for robot learning and operation, these strengths will further accelerate LG's transformation into a comprehensive robotics solutions provider with both hardware and software capabilities. "Through the synergy built on 'One LG' - bringing together core capabilities across the Group - and strategic collaboration with global partners, we will secure our competitiveness in physical AI and become a comprehensive robotics solutions provider," said Lyu Jae-cheol, CEO of LG Electronics. About LG Electronics, Inc. LG Electronics is a global innovator in technology and consumer electronics with a presence in almost every country and an international workforce of more than 75,000. LG's four Companies - Home Appliance Solution, Media Entertainment Solution, Vehicle Solution and Eco Solution - combined for global revenue of over KRW 89 trillion in 2025. LG is a leading manufacturer of consumer and commercial products ranging from TVs, home appliances, air solutions, monitors, automotive components and solutions, and its premium LG SIGNATURE and intelligent LG ThinQ brands are familiar names world over. Visit www.LG.com/global/newsroom/ for the latest news.

All News North Korea Sports Top News Most Viewed Korean Newspaper Headlines Today in Korean History Yonhap News Summary Editorials from Korean Dailies URL is copied. This will let Google show Yonhap news articles that match or are related to your search SEOUL, Aug. 18 (Yonhap) -- Madison Huang, a senior director at Nvidia Corp. and the daughter of Chief Executive Officer (CEO) Jensen Huang, visited LG Electronics Inc.'s robotics hub in South Korea on Tuesday. The younger Huang met with senior LG Electronics officials to discuss ways to expand their partnership in artificial intelligence (AI) infrastructure and robotics. The visit comes after LG Electronics announced last week that it plans to unveil a next-generation humanoid robot capable of bipedal walking in the first quarter of 2027, based on Nvidia's robotics platform. The South Korean tech giant signed a memorandum of understanding with Nvidia at the U.S. chipmaker's headquarters in Santa Clara, California, with LG Group Chairman Koo Kwang-mo and Jensen Huang in attendance. During Tuesday's visit, representatives from the two companies reviewed the operation of LG Electronics' robotics factory in Seoul, which is being built with the goal of becoming fully operational by the end of this year. The facility currently uses LG Electronics' proprietary humanoid robot, LG CLOiD, to generate, collect and process data for AI training. Nvidia's physical AI technologies, including the Nvidia Omniverse libraries, Cosmos world foundation models and Isaac open robotics development platform, are being used in the data collection and application process. Madison Huang, a senior director at Nvidia Corp. and the daughter of Chief Executive Officer (CEO) Jensen Huang, visits LG Electronics Inc.'s robotics hub in South Korea on Aug. 18, 2026, in this photo provided by LG Electronics. (PHOTO NOT FOR SALE) (Yonhap) khj@yna.co.kr(END) All News National North Korea Economy/Finance Biz Culture/K-pop Sports Images Videos Top News Most Viewed Korean Newspaper Headlines Today in Korean History Yonhap News Summary Editorials from Korean Dailies Korea in Brief Useful Links Weather Advertise with Yonhap News Agency Korea Annual

A portable, high-fidelity robot-free data-production system for robot manipulation learning. Across three policy backbones evaluated in the report, policies post-trained solely on HiFi-UMI data reached success rates comparable to policies post-trained on in-domain real-robot teleoperation data. NEW YORK, Aug. 20, 2026 /PRNewswire/ — Simple AI has published the Tech Report for HiFi-UMI, a high-fidelity robot-free data-production system for robot manipulation learning, together with HiFi-UMI-2K, a 2,000-hour open dataset released under the Creative Commons Attribution 4.0 license. The Tech Report is available on arXiv and the dataset on Hugging Face. A portable, high-fidelity robot-free data-production system for robot manipulation learning. Across three policy backbones evaluated in the report, policies post-trained solely on HiFi-UMI data reached success rates comparable to policies post-trained on in-domain real-robot teleoperation data. Progress in robot manipulation learning is increasingly constrained by data. Real-robot teleoperation yields accurate, directly trainable trajectories but is difficult to scale: every hour of data requires the target robot, a teleoperation rig, and a skilled operator. Robot-free handheld demonstrations are cheaper and easier to scale, but have primarily been used for pre-training. Task-specific post-training, the stage that grounds a policy for real-robot deployment, has typically continued to rely on a smaller amount of real-robot teleoperation data as an anchor. The report examines whether raising the fidelity of robot-free demonstration data, rather than shrinking the real-robot fraction, can remove that anchor for target-task post-training. HiFi-UMI is a portable data-production system co-designed for four fidelity properties. Pose accuracy comes from head-mounted offline stereo-inertial SLAM, which the report measures at 3mm workspace-local end-effector accuracy. Cross-sensor timing is aligned to below 40 microseconds through a shared hardware trigger across all cameras and sensors. Inter-gripper relative pose is measured natively rather than reconstructed. Per-hand sensing covers approximately 200 degrees of field of view through two non-parallel wide-angle fisheye cameras. Every captured demonstration passes through automatic trajectory reconstruction and simulation replay validation, each gate with an approximately 98% pass rate. The report evaluates the approach across three policy backbones spanning the vision-language-action and world-action-model families, and four bimanual tabletop tasks. Policies post-trained solely on HiFi-UMI demonstrations reached success rates comparable to policies post-trained on in-domain real-robot teleoperation data, with reported differences of −2.5, +3.1, and −0.6 percentage points across the three backbones. On a precision insertion task, the strongest HiFi-UMI-only policy reached 85% success under conditions where the teleoperation baseline had the advantage of being collected in the evaluation scene. Separately, pre-training on 4,000 hours of the same corpus reduced offline action prediction error on ten unseen tasks by 41%, and increased real-robot success on one of the evaluated backbones by 18.1 percentage points at matched post-training data. “We wanted to test whether fidelity, rather than scale alone, is what unlocks robot-free data for deployment-oriented training,” said Xiaofei Li, founder of Simple AI. “The report shows what this can look like within a specific set of tasks and models. By open-sourcing HiFi-UMI-2K, we hope to give the wider research community a shared, high-fidelity resource for continuing to study this question.” The report characterizes these findings as approximate aggregate parity within the tested models, tasks, and experimental conditions. The deployment robot uses the same gripper and wrist-camera configuration as the capture setup, with the main embodiment difference being robot arm kinematics. The comparison is not sample-matched, with 3,200 HiFi-UMI trajectories per task set against approximately 300 teleoperation trajectories, and reflects a comparison between practical data-production pipelines rather than a claim of per-trajectory equivalence. The report does not generalize the result to all robot learning settings, and does not conclude that real-robot data is no longer required in the broader field. HiFi-UMI-2K is distributed in a training-ready format with synchronized multi-view video, bimanual end-effector trajectories, gripper states, language annotations, and subtask boundaries. Human faces in the recordings are masked before release. The paper reached No. 1 on Hugging Face Daily Papers on July 29. HiFi-UMI is one component of Simple AI’s work across foundation models, high-fidelity data, robotic systems, and real-world deployment. The company welcomes conversations with research groups and industry partners interested in high-fidelity data for robot learning. About Simple AISimple AI is an embodied AI company developing general-purpose embodied intelligence systems for human living spaces. Its work integrates foundation models, high-fidelity data, robotic systems, and real-world deployment across the full embodied AI stack. Resources Tech Report: arxiv.org/abs/2607.25895 Dataset: huggingface.co/datasets/simple-world-lab/HiFi-UMI-2K Project page: cloud.simpleai.tech/simple-world-lab/hifi-umi/ Media Contact Grant Xin Simple AI [email protected] ​ Your email address will not be published. Required fields are marked * Comment * Name * Email * Website Save my name, email, and website in this browser for the next time I comment. Copyright © 2024 Money Compass Media (M) Sdn Bhd. All Rights Reserved Login to your account below Remember Me Please enter your username or email address to reset your password. Copyright © 2024 Money Compass Media (M) Sdn Bhd. All Rights Reserved

From daily news and career tips to monthly insights on AI, sustainability, software, and more—pick what matters and get it in your inbox. Access expert insights, exclusive content, and a deeper dive into engineering and innovation all with fewer ads or a completely ad-free experience. All Rights Reserved, IE Media, Inc. Follow Us On Access expert insights, exclusive content, and a deeper dive into engineering and innovation all with fewer ads or a completely ad-free experience. All Rights Reserved, IE Media, Inc. Humanoids promise flexibility, but industrial robots already dominate factories. The humanoid robot has become the new face of factory automation. It can walk between workstations, pick up components, and potentially operate equipment originally designed for people. Yet, away from the carefully edited demonstrations, conventional industrial robots continue to perform most automated factory work. This raises a less glamorous but more important question: Will factories be transformed by human-shaped machines, or by increasingly capable versions of the robots they already use? The answer depends on what manufacturers need. Humanoids promise flexibility, but industrial robots offer something factories value even more: proven productivity. Industrial robots do not need to wait for their breakthrough moment. It happened decades ago. Factories installed 542,000 industrial robots in 2024, while the number operating worldwide reached approximately 4.66 million, according to the International Federation of Robotics. Annual installations have exceeded 500,000 for four consecutive years. That installed base has created an ecosystem of manufacturers, software developers, component suppliers, safety specialists and maintenance technicians. Companies know how to calculate the return on a robotic arm used for welding, painting, palletizing or assembly. Purpose-built robots also have a mechanical advantage. A fixed arm does not need to walk, balance, or understand an entire factory. It only has to perform a defined movement quickly and accurately, sometimes thousands of times a day. When a production task rarely changes, specialization beats versatility. A machine optimized for one operation will usually be faster, more reliable, and easier to maintain than a general-purpose humanoid. Collaborative robots are also weakening one of the main arguments for humanoids. These machines can work closer to people and be reprogrammed for different tasks. Cobots represented 10.5% of industrial robot installations in 2023, the IFR reported. Humanoids, therefore, are not competing against yesterday’s factory robots. They are competing against an established technology that is becoming safer and more adaptable. Industrial robots are effective when factories can organize work around them. The calculation changes when manufacturers cannot justify rebuilding a production line for one difficult or irregular task. Factories are filled with stairs, carts, shelves, tools and workstations designed for the human body. A humanoid could enter these environments without requiring new conveyors, protective cages or major structural changes. The same robot might also perform several jobs. It could deliver components, tend a machine and inspect parts at different points during a shift. If manufacturers can add abilities through software, one hardware platform could eventually replace several narrowly designed systems. Early deployments show where this opportunity may emerge. Mercedes-Benz is testing Apptronik’s Apollo humanoid for moving components and carrying out initial quality checks. The company is initially concentrating on repetitive intralogistics work rather than core vehicle assembly. In logistics, GXO signed a multiyear agreement with Agility Robotics to deploy Digit. At a Spanx facility, the humanoid moves totes from mobile robots and places them onto conveyors. These are real jobs, but they also reveal the current limits of the technology. Humanoids are beginning with controlled, repetitive material-handling tasks—not roaming factories and independently switching between dozens of operations. A humanoid’s humanlike design is useful only when the task requires it. Two legs consume energy and introduce the risk of falling. Dexterous hands add mechanical complexity, while autonomous software must remain dependable throughout production. Battery life presents another hurdle. Agility Robotics said in 2025 that upgrades had extended Digit’s operating time to up to four hours. Continuous factory work may still require charging, battery swaps or multiple robots. Factories ultimately measure machines through output, uptime, safety and cost – not how closely they resemble workers. For that reason, industrial robots will transform factories first. They already operate at scale and remain the better choice for stable, high-volume processes. Humanoids will not sweep them aside. Their opportunity is to automate the spaces between conventional robotic cells: jobs that require mobility, adaptability, and interaction with human-designed surroundings. The future factory will likely use both. Fixed robots will dominate repetitive production, mobile robots will transport goods, cobots will share workstations with people, and humanoids may tackle the stubborn tasks that earlier forms of automation could not reach. Atharva is a full-time content writer with a post-graduate degree in media & amp; entertainment and a graduate degree in electronics & telecommunications. He has written in the sports and technology domains respectively. In his leisure time, Atharva loves learning about digital marketing and watching soccer matches. His main goal behind joining Interesting Engineering is to learn more about how the recent technological advancements are helping human beings on both societal and individual levels in their daily lives. Premium Follow

Get the best South African technology news and analysis delivered to your e-mail inbox every morning. The CEO of Chinese humanoid robot unicorn Unitree said on Thursday the industry is edging towards a “ChatGPT moment” for robot brains, after the company’s blockbuster Shanghai listing crystallised China’s ambitions to lead the next frontier of AI-powered machines. Shares in China’s best-known humanoid robot maker fell 11% on Thursday, a day after soaring nearly sixfold in their market debut, highlighting intense investor enthusiasm for a fledgling sector that enjoys strong backing from Beijing. “We are marching towards a ‘ChatGPT moment’ in embodied intelligence,” said Wang Xingxing, the Hangzhou-based start-up’s founder and CEO, at a major robot conference in Beijing. The global AI boom that has reshaped the world economy was sparked in late 2022 by ChatGPT’s breakthrough large language model, a watershed moment that drove mass adoption. No comparable inflection point has yet emerged for world models, the physical AI simulation systems designed to help robots understand and navigate real-world environments. Wang said the industry is nearing a breakthrough where robots can be placed in unfamiliar environments and complete most tasks through simple voice or text instructions. “We hope that in the future, we can see a robot be introduced into an unfamiliar household and it can achieve approximately 80% of tasks successfully through voice or text commands,” he said. “It is an important tipping point for the robot industry to usher in explosive growth.” At the same time, Wang cautioned that a major leap in robot software could arrive within two to three years in an optimistic scenario, or within five to 10 years at the latest. “Relative to the mood around World Robot Conference and Unitree’s spectacular IPO, Wang Xingxing was notably sober about current capabilities,” said Georg Stieler, head of automation at robotics consultancy Stieler. Unitree is the world’s largest producer of robot dogs and the second biggest maker of humanoid robots by shipments, according to industry data. It rose to fame with impressively choreographed performances of robots dancing and performing kung fu, showcased on Chinese television, and is increasingly deploying its machines in real-world industrial settings. Its IPO prospectus shows most customers are universities and research institutions. Wang He, founder of Chinese robotic start-up Galbot, expects the sector’s “ChatGPT moment” to arrive by 2028, defining it as the point when robots can perform about 70-80% of everyday tasks without specialised training. “With continued accumulation of data and further technological breakthroughs, we expect to reach the ‘ChatGPT moment’ for embodied intelligence by 2028,” he said. Unitree’s Wang said the company’s biggest current investment in terms of capital and manpower is in world models and that it is “lagging behind” in the real-world application of physical AI models. He also said humanoids are not yet capable enough for mass deployment, pointing to limitations in the AI models that power robots’ decision-making and interactions as the industry’s biggest bottleneck. Even so, Wang expects the sector to enter a decade of rapid autonomous evolution of humanoid robots, as the AI boom accelerates and AI language models learn to improve themselves. China delivered over 40 000 humanoids in the first half of this year alone, and accounts for 97% of global shipments of humanoid robots, a Chinese humanoid industry body said in a report Thursday. Beijing is betting that robots can eventually replace human labour in repetitive, low-value and dangerous settings as it faces a shrinking workforce due to demographic decline. While humanoids are gradually being introduced to factory floors and logistics warehouses, they remain less efficient than human workers in most applications. Humanoid robots have also become a new front in US-China technology rivalry. Last month, the US Federal Communications Commission banned future imports of foreign-made humanoid and quadruped robots citing national security concerns, in a major blow to Chinese producers. Several companies at the World Robot Conference in Beijing said they were looking to expand overseas. — Ju-min Park, Eduardo Baptista and Laurie Chen, (c) 2026 Reuters Get the best South African technology news and analysis delivered to your e-mail inbox every morning. Type above and press Enter to search. Press Esc to cancel. TechCentral uses cookies to enhance its offerings. Consenting to these technologies allows us to serve you better. Not consenting or withdrawing consent may adversely affect certain features and functions of the website.

Listen to this article in summarized format (Catch all the Technology News News, and Latest News Updates on The Economic Times.) ...more Popular Categories Hot on Web In Case you missed it Top Searched Companies Other useful Links Top Calculators Top Slideshow Top Story Listing Top Prime Articles Top Definitions Top Commodities Private Companies Top Market Pages Latest News follow us on Download ET App:

Company’s robots have gained global fame via videos of them performing martial arts and as dancers for pop stars Unitree, the world’s biggest humanoid robot maker, has made a spectacular entry on to China’s stock market, with its shares surging by more than 600%. The Chinese company’s robots have gained global fame via viral videos of them performing martial arts, running at Olympic speeds and serving as backup dancers for pop stars. Shares in the business, officially known as Yushu Technology Co, rose to as high as 1,100 yuan (£120.39) on Wednesday, up from an IPO price of just 150.8 yuan. Its gains were later pared back to a rise of nearly 500%. Investors are searching for winners in robotics development, which has emerged as one of the key battlegrounds in the AI race. Unitree, which was founded in 2016, shipped more than 5,500 humanoid robots last year. The market for human-like robots is expected to grow rapidly, with analysts projecting that sales could rise from around $2bn (£1.5bn) in 2025 to $300bn by 2035. There was exceptional demand from Chinese retail investors in Unitree’s IPO, with the tranche of shares dedicated to non-professional stockpickers oversubscribed by thousands. Unitree is one of the few listed humanoid robot makers in the world. Its biggest competitor, AgiBot, is private and its smaller rival UBTech is listed in Hong Kong. However, at least half a dozen other Chinese humanoid robotic businesses are preparing to go public, including Deep Robotics and Leju Robotics. Wang Xingxing, who founded Unitree in 2016 and remains its chief executive, owns about a fifth of the business. The spike in its share price means his personal wealth is now worth more than $12bn on paper, according to Reuters. Unitree’s market debut also coincided with the opening of the World Robot Conference in Bejiing on Wednesday, where hundreds of companies, mostly Chinese, will launch new products and demonstrate their technical developments. Sign up to Business Today Get set for the working day – we'll point you to all the business news and analysis you need every morning after newsletter promotion Last month, the US Federal Communications Commission banned imports of future models of foreign-made humanoid and quadruped robots based on national security concerns. This summer the Pentagon also added Unitree to a list of Chinese military companies, describing it as a “contributor to the Chinese defence industrial base”. Unitree has previously said its robots are for civilian use. The business is backed by several big Chinese technology companies, including Tencent and Alibaba.

Listen to this article in summarized format (Catch all the Technology News News, and Latest News Updates on The Economic Times.) ...more Popular Categories Hot on Web In Case you missed it Top Searched Companies Other useful Links Top Calculators Top Slideshow Top Story Listing Top Prime Articles Top Definitions Top Commodities Private Companies Top Market Pages Latest News follow us on Download ET App:

Winnipeg 24° C, Partly cloudy Full Forecast © 2026 Winnipeg Free Press Quick Links Ways to support us Replica E-Edition Business Arts & Life Sports Opinion Media Homes Canstar Community news About Us This browser doesn't support push notifications at the moment. Check browsers features, update your browser or try to use one from the list of recommended to manage your notifications settings: If you wish to manage your notification settings from this browser you will need to update your browser's settings for this site. Just click button below and allow notifications for this site Note Safari 16.4+ working on iOS devices also need this site app to be installed at device's Home Screen for Push Notifications to work Notifications are blocked for this site. If you wish to manage your notification settings from this browser you will need to update your browser's settings. Usually you'd need to click on site options icon to the left of address bar and change notifications preferences/permissions from there Urgent and important stories Noteworthy news and features Advertisement Advertise with us HONG KONG (AP) — Shares of Unitree, one of China's largest humanoid robot makers, initially soared as much as 629% in its public stock trading debut Wednesday in Shanghai, in the latest highlight of investor optimism over China’s technological advances. Read this article for free: Already have an account? Log in here » To continue reading, please subscribe: One year of digital access for only $205* *First annual payment billed as $205.00 + GST for one year. This annual subscription will automatically renew at $233.00 + GST every 52 weeks (10% off the regular annual price of $259.35). Offer available to new and qualified returning subscribers only. Cancel any time. To continue reading, please subscribe: $1 for the first 4 weeks* *Your next Brandon Sun subscription payment will increase by $1.00 and you will be charged $17.95 plus GST for four weeks. After four weeks, your payment will increase to $24.95 plus GST every four weeks. HONG KONG (AP) — Shares of Unitree, one of China's largest humanoid robot makers, initially soared as much as 629% in its public stock trading debut Wednesday in Shanghai, in the latest highlight of investor optimism over China’s technological advances. Read unlimited articles for free today: Already have an account? Log in here » HONG KONG (AP) — Shares of Unitree, one of China’s largest humanoid robot makers, initially soared as much as 629% in its public stock trading debut Wednesday in Shanghai, in the latest highlight of investor optimism over China’s technological advances. The company, founded in 2016 by entrepreneur Wang Xingxing in the eastern Chinese technology hub of Hangzhou, raised around 6.1 billion yuan ($904 million) in its listing on the Shanghai Stock Exchange’s Nasdaq-style STAR market. The shares were priced at 150.80 yuan ($22.36) a share. Its shares closed 460% higher at 845 yuan ($125.31). Advanced robotics, often referred to as “embodied” artificial intelligence, is a key sector in the heated technology rivalry between China and the U.S. China leads the U.S. in terms of production capacity of humanoid robots and the ability to scale up manufacturing. Last year, of the roughly 15,000 humanoid robots shipped globally, Unitree and AGIBOT, another major Chinese humanoid robot maker, each shipped more than 5,000, according to technology research and advisory group Omdia, way ahead of their U.S. counterparts. For the first half of 2026, Omdia’s estimates put Chinese humanoid robot makers’ total global shipments at around 18,500 units. But analysts say for now, many of these humanoid robots are still mainly used for demonstrations, performances and research rather than real-world applications. Unitree’s humanoid robots, for example, dazzled audiences at China’s annual Spring Festival gala by performing backflips and martial arts. “The real competitive test will be whether companies – Chinese or American – can achieve reliable performance and attractive returns on investment in large-scale industrial and commercial deployments,” analyst Kangyuxiao Li at investment research firm Morningstar said in a commentary. Unitree said proceeds from the IPO will be used for advanced robot research and development and development of its robot manufacturing base. The company is the first publicly traded humanoid robotics maker in mainland China. Analysts believe the IPO could set a precedent for valuations of other robotic company stock offerings. It “gives mainland investors direct exposure to one of the sector’s leading companies,” said Li at Morningstar. Unitree’s debut coincided with the World Robot Conference in Beijing, featuring more than 300 exhibitors. Chinese leaders made robots a development priority. UBTech, another major Chinese humanoid robot maker, has shares traded in Hong Kong. Its shares fell more than 10% on Wednesday. Unitree reported about 1.7 billion yuan (about $250 million) in revenue in 2025, mainly from sales of its humanoid robots and quadruped robots, often referring to four-legged robot dogs. More than 40% of that revenue was from overseas. Last year, the U.S. accounted for roughly 13% of Unitree’s revenue, the company said. But that was before the U.S. Federal Communications Commission banned imports of new foreign-made humanoid and quadruped robots in July on national security grounds. Unitree cautioned that the U.S. ban, which it said applies to its new models, could affect its future sales in the U.S. market. Existing models of its advanced robots can still be sold, but Unitree faces the risk that the U.S. could further expand its restrictions. Lian Jye Su of Omdia said Chinese robot makers will likely seek to expand in major markets outside of the U.S., such as Europe. Advertisement Advertise With Us Advertisement Advertise With Us Advertisement Learn more about Winnipeg Free Press Advertising solutions RCMP are seeking the public’s help to identify a person of interest after an all-terrain vehicle and other belongings were stolen from a cottage in southern Manitoba. The theft happened near near Provincial Road 440 in the Rural Municipality of Louise, about 150 kilometres southwest of Winnipeg, on Aug. 11. Police said surveillance video showed a man walking onto the property at about 2:30 p.m. while covering his face and holding power tools. A grinder was used to open a locked shipping container, where a 1982 Honda ATC200 all-terrain vehicle, a Honda EU2000i generator and a Stihl FS90 trimmer were among the items stolen, RCMP said. One of two Spypoint trail cameras was also reported stolen. OTTAWA - Prime Minister Mark Carney is scheduled to brief his cabinet and Canada's premiers today about the potential trade deal reached with the United States on Tuesday. Carney will chair a virtual cabinet meeting and then hold a virtual meeting with premiers to discuss the deal that appears to have averted a new round of punishing tariffs on Canada. President Donald Trump paused for three days the latest tariff threat on $28 billion worth of Canadian goods pending the finalization of the deal. The U.S. said the threat was in response to trade irritants including the provincial bans on American alcohol in most provinces and Canada's dairy supply management system. To solve our puzzles, please subscribe with this special offer: | Two men in electric ankle bracelets who were being monitored by justice officials have been accused in a two-month spree of commercial break-ins, vehicle thefts and other property crimes. City police have charged one of the suspects and are searching for the second. About $140,000 worth of property, including trailers, lawn equipment, motorcycles and bicycles, was reported stolen in 13 incidents in Winnipeg and the Rural Municipality of Headingley between April 20 and June 28, the Winnipeg Police Service said. Both suspects were wearing electronic monitoring bracelets during that time. In the Second World War, a grand coalition, held together politically by U.S. President Franklin D. Roosevelt, succeeded in defeating Nazi Germany, after many bumps in the road. The rich history of this period reinforces the truism that coalitions — bringing together parties of divergent strength and interests to face together a common challenge — are fraught with many points of potential fracture. Canada was not only a partner in the great international war effort but also fashioned a national coalition — presided over by Prime Minister Mackenzie King — necessitated by the federalist nature of our government. It worked after a fashion — conscription notwithstanding — and we emerged as a trusted friend and ally of our southern neighbour. Oh, how times have changed. We are again faced with an existential threat; not to be compared to a world war, but serious enough given it’s posed not by a distant foe on another continent, but by the most powerful country in the world, with whom we share 8,800 kilometres of heretofore undefended border. The unprovoked attack on Canada has been economic; tariffs, threats of tariffs, scrapped trade agreements and so forth — measures that have already had significant impact and promise to escalate. How does a country of 10 quasi-independent provinces with a national government constitutionally pre-eminent in international relations respond? An internal investigation by Manitoba’s Families Department is nearly complete in the case of a six-year-old girl who is paralyzed after being assaulted by her foster mother in Winnipeg. The review by child-welfare officials is intended to prevent similar incidents, but it’s unlikely the public will learn the findings or recommendations. “We’re working closely with the (child and family services) authority to make sure that every recommendation is taken seriously and implemented,” Families Minister Nahanni Fontaine said Monday. “They’re important (investigations) because they help us identify what went wrong so we can make those changes to ensure something like this tragedy never happens again.”

(ECNS) -- Chinese robotics company Unitree Robotics unveiled a new humanoid robot dubbed “Superman” on Monday, saying it can perform a 2-meter jump from a standing position and reach a top speed of 12.66 meters per second with 0.85-meter-long legs. The company said the robot surpasses human records for both vertical jump and sprint speed. Developed in just over three months, the prototype will continue to be refined in the coming months. Unitree also introduced the As2W, a wheeled-leg quadruped robot designed for heavy loads and long-duration tasks. It can carry up to 16 kilograms continuously, travel more than 30 kilometers unloaded, and reach speeds of 6 meters per second. Demonstrations showed the robot diving from platforms and walking through shallow water with its IP54-rated design. The 25-kilogram machine has a maximum payload of 180 kilograms. On May 12, Unitree unveiled the GD01, its first mass-produced manned transforming mecha, with a starting price of 3.9 million yuan (about $540,000). It is the most expensive publicly priced product released by the company to date. (By Gong Weiwei) Over 2,000 robots from 16 countries to compete at World Humanoid Robot Games Residents enjoy AI-powered games at robot expo in Yinchuan World Robot Contest Jiangsu qualifier kicks off in Yangzhou

news Alerts There are no new alerts at this time Chinese robotics company Unitree unveiled a new humanoid robot dubbed “Superman” which it says is “breaking the limits of humanity.” According to the company, the robot can perform a vertical jump of over 6 feet and can reach a top speed of more than 40 feet per second.Aug. 19, 2026 © 2026 NBCUniversal Media, LLC

Switch between CA and NY editions here. See more of our coverage in your search results. Leading Chinese robotics firm Unitree – known for its backflipping humanoid robots and robotic dogs – surged 460% on Wednesday after making its public debut in Shanghai. Unitree – whose dexterous robots have gone viral with displays of dancing and martial arts, while raising national security concerns in Congress – raised 6.1 billion yuan, or about $904 million, in its public debut. The Hangzhou-based firm is the first robotics firm from mainland China to go public, with investors including AI firm DeepSeek and tech giant Tencent. The monster public debut came just two days after Unitree unveiled a next-gen humanoid robot it calls “Superman,” which is purportedly capable of leaping more than six feet in the air and running at more than 28 miles per hour. “The mass-production inflection point appears close,” JPMorgan Chase & Co. analysts said in a note, according to Bloomberg. Wang Xingxing, Unitree’s 36-year-old CEO, saw his personal wealth surge as high as $16 billion during the blockbuster trading debut, according to Forbes. The company generated about $250 million in revenue last year, with about 13% of that total coming from the US, according to disclosures. In the US, Elon Musk’s Tesla has unveiled plans to build huge numbers of “Optimus” humanoid robots, but they have yet to enter mass production. As The Post reported back in 2024, US officials have grown increasingly concerned about the rapid progress of humanoid robots made by Unitree and other China-based firms. Some have voiced fears that the Chinese Communist Party or state-sponsored bad actors could use them to sabotage critical infrastructure or spy. Last month, the FCC banned imports of foreign-made humanoid robots and “robot dogs” in what was widely seen as a measure targeting China, which controls about 85% of the global market. FCC Chairman Brendan Carr said the ban, which only applies to “new versions” of the robots, was meant to “secure America’s critical supply chains.” Elsewhere, the Pentagon in June added Unitree to a list of companies that work with the Chinese military. With Post wires