How humanoids learn to read the room
- RBR50 Winners 2023 ## Conclusion As humanoids move into more complex roles, the safety, sensing and interaction demands will only grow. Ostrem believes the future lies in better AI at the edge — both in terms of improved accuracy in object classification...
Key takeaways
- Figure AI has dramatically accelerated production of its third‑generation Figure 03 humanoid, moving from one robot per day to one per hour in just 120 days and reporting over 350 units built and an 80 % first‑pass yield, while its “System 0” controller now fuses visual and proprioceptive data to let robots navigate stairs and uneven terrain without real‑world fine‑tuning.
- NVIDIA unveiled the Isaac GR00T reference humanoid, built on a Unitree H2 Plus body, Sharpa five‑fingered hands and the Jetson Thor compute module, offering an open hardware and software stack for academic research.
- NVIDIA also partnered with Unitree to market a research‑focused humanoid platform that integrates the new Blackwell chip, with sales slated for later in 2026 and early adopters such as Stanford and ETH Zurich.
- At the Humanoids Summit in Tokyo (May 28 2026), Japanese and Chinese developers showcased advanced dexterous hands capable of threading needles, child‑like dancing robots, and adult‑size delivery bots, highlighting China’s rapid progress with companies like Booster Robotics, LimX Dynamics and Unitree supplying many of the showcased mechanisms.
- Meanwhile, industry leaders such as BMW, Toyota, Hyundai and Boston Dynamics are testing or deploying humanoid robots—including Agility Robotics’ Digit and Boston Dynamics’ Atlas—in automotive factories, emphasizing fleet‑management tools, OTA updates and fault‑diagnostics to integrate robots alongside human workers.
- RBR50 Winners 2023 ## Conclusion
As humanoids move into more complex roles, the safety, sensing and interaction demands will only grow. Ostrem believes the future lies in better AI at the edge — both in terms of improved accuracy in object classification, and optimized low-power edge processing.
ADI already has sensing and perception, connectivity and battery management figured out in the automotive sphere. The natural next step is leveraging those technologies into emerging applications like humanoids.
“In some ways, humanoid robots are where cars were many years ago,” says Ostrem. “The architecture isn’t fully set, and there is significant room for industry collaboration around standardizing interfaces to stimulate the ecosystem around humanoid robots.” Already in wide use throughout the automotive industry, ADI’s Gigabit Multimedia Serial Link (GMSL) technology transports video data in real time, in a single stream capable of carrying many gigabits per second. In humanoid robots, this enables redundancy and fast, local processing of visual data, which enables these systems to identify and understand what the visual data means about their surroundings using physical AI to process visual data locally in the robot, rather than sending it to the cloud.
Audio
Vision alone is not enough, however; if a robot is going to work collaboratively with humans, it needs intelligent hearing, too. “Being able to speak in natural language and having a conversational user interface is a very powerful way to communicate with a robot,” says Ostrem.
Mentioned in this article