Physical AI, humanoid robots, robot learning, and embodied intelligence from Figure, Boston Dynamics, Tesla, and research labs.
439 articles
A fresh benchmark result showing a 56% score for Figure's humanoid robot is prompting deeper scrutiny of how well embodied AI agents actually generalize across tasks. The finding is pushing companies like Unitree and Agibot to rethink strategies for proving real-world adaptability rather than narrow task performance. Source: eu.36kr.com
Amid the ongoing robotics investment boom, one industry expert points to three US-based humanoid robot companies worth watching. The pick reflects growing investor interest in translating robotics hype into concrete stock opportunities. Source: forbes.com
A Chinese startup developing AI systems for humanoid robots claims that by next year, robots could complete most general-purpose tasks based purely on verbal commands. However, the company acknowledges that true everyday usefulness in homes and workplaces will still take longer to achieve. Source: reuters.com
Industry analysts point to 2026 as a pivotal year for Physical AI, where edge computing enables robots and industrial systems to process data locally in real time. Engineers and developers are being urged to understand how this shift affects design choices in robotics and automation hardware. Source: embeddedcomputing.com
Boden AI has open-sourced the RW-RL-HIL-Dataset, containing over 82 hours of real-world robot reinforcement learning data combined with human intervention records. The release, announced through Gasgoo Embodied Intelligence, aims to support development of more robust real-world robotic RL systems. Source: autonews.gasgoo.com
Figure has introduced Helix 2.5, described as the company's most advanced neural network to date. The model reportedly achieves zero-shot generalization across 30 different home environments. Source: figure.ai
Harvard researchers have launched RLE-Bench, a benchmark designed to evaluate how well AI coding agents can handle the engineering of physical robotic systems. The tool aims to measure real-world engineering capability beyond typical software tasks. Source: seas.harvard.edu
Arm's Total Design program for Physical AI has brought together more than 80 developers to address shared challenges in autonomous vehicles and robotics. The initiative focuses on common ground in perception, AI processing, real-time control, safety, and computing efficiency. Source: therobotreport.com
Goldman Sachs forecasts that 6.5 million humanoid robots will be in use by 2035, marking a shift from traditional automation toward physical AI. Companies including Amazon, Tesla, and NVIDIA are pushing robotics forward across multiple industries. Source: streetwisereports.com
At its yearly 919 launch event, Faraday Future plans to introduce nine new AI-powered robots along with four industry-specific solutions. The reveal is meant to complete the company's 'one-brain, multi-form, multi-capability' robot platform, dubbed Robot World 2.0. Source: morningstar.com
Boston Dynamics has released Spot 5.2, which uses AI, cameras and sensors to autonomously detect and respond to issues in industrial settings. The update marks another step in Hyundai's broader strategy to bring physical AI into real-world operations. Source: upi.com
As AI expands beyond digital interfaces into physical environments, organizations face new questions about where it can generate the greatest business value. Industry leaders are being urged to identify these overlooked opportunities before competitors do. Source: mckinsey.com
Dermot O'Driscoll, Arm's VP of go-to-market for physical AI, will present core compute principles needed to build trustworthy physical AI systems. The talk is expected to address how hardware design underpins reliable real-world AI deployment. Source: therobotreport.com
China's data authority announced plans to create standards for training data used in embodied AI, responding quickly to an industry request made just over a week earlier. By comparison, the EU's Data Act regulates machine-generated data that hasn't even started being collected yet. Source: thenextweb.com
Goldman Sachs released a new forecast projecting a much larger humanoid robotics market than previously expected, sending ripples through chip and robotics-related stocks. The bank's updated outlook suggests the industry could scale far faster than earlier estimates indicated. Source: 247wallst.com
Industry voices argue that no single robot can match the effectiveness of well-coordinated multi-robot teamwork, shifting focus toward swarm intelligence as the sector's next big breakthrough. This collective approach is increasingly seen as more realistic than pursuing one universally capable humanoid robot. Source: eu.36kr.com
Machines have already become part of everyday human life, so the real debate now is about what form that integration should take next. The article explores how humanoid robots and cyborg technologies are reshaping the boundary between human and machine. Source: forbes.com
China's National Data Administration met with companies including JD.com and the Beijing Academy to discuss embodied AI, signaling government support for data infrastructure in the sector. The meeting also focused on establishing common standards for training data used by embodied AI systems. Source: mlex.com
China's data watchdog announced plans to draft standards for embodied AI and to guide local governments in implementing related policies. The move comes as demand for physical AI systems, such as robots, continues to grow. Source: bloomberg.com
Mecka AI, a startup that gathers and analyzes human movement data to train humanoid robots and other robotics systems, is reportedly closing in on a new funding round led by Sequoia Capital. The deal reflects growing investor interest in companies supplying high-quality motion data for robotics AI. Source: techcrunch.com