Tag
The paper proposes DissipNet, a deep discrete-time dissipative recurrent neural network that explicitly enforces dissipativity through structural constraints to ensure stable modeling of dissipative systems, outperforming traditional RNNs and Physics-Informed Neural Networks.
Microsoft Research findings demonstrate that offloading AI inference from robots to edge or cloud systems enhances task success rates, efficiency, and battery life for physical AI applications.
PhysAI-Bench is a benchmark for evaluating LLM-based agentic decision-making in autonomous UAV systems, comprising over 10,000 standardized instances and evaluating 29 foundation models.
Aaron Edsinger demonstrates Hello Robot's Stretch 4 home-assistance robot at TechCrunch Disrupt 2026, showcasing its practical features for users with mobility impairments.
The article emphasizes the necessity of comprehensive safety measures for deploying Physical AI at scale, such as in autonomous vehicles and robots, and introduces NVIDIA Halos as a full-stack safety system to address these challenges.
Vantora, a startup builder formerly known as UP.Labs, raised $100 million from Silversmith Capital Partners and is shifting its focus to building proprietary physical AI startups for corporate customers.
Nvidia's Les Karpas will speak at TechCrunch Disrupt 2026 about the challenges and opportunities in robotics awaiting a breakthrough moment similar to ChatGPT.
World models for robotics enable faster and cheaper training by generating simulated footage, exemplified by LTX-2.5, to achieve advanced physical navigation in machines.
This paper explores building self-adaptive physical AI agents using LLMs to manage long-horizon tasks in a zero-shot manner, showing they can adapt effectively to environmental changes compared to reinforcement learning agents.
OpenArm is an open-source 7DOF humanoid arm designed for physical AI research, featuring high backdrivability, compliance, and practical payloads to enable safe human-robot interaction and reproducible experimental conditions.
The article highlights the rapid progress in humanoid robotics from 2015 to 2026, showcasing advanced capabilities like running and industrial tasks, and speculates on future developments by 2036.
This paper identifies and calibrates seven non-exclusive sources of capability formation in Physical AI, such as recorded experience and embodied coupling, using a reconstructive inductive design to establish a falsifiable vocabulary for capability analysis.
Skild AI announced reaching $100M ARR in just 10 months after starting deployments of its physical AI robots in various industries, underscoring the growing impact of AI in transforming large-scale sectors like manufacturing and logistics.
Skild AI has launched the S1 robot foundation model, which enables robots to learn new tasks from a single video demonstration using NVIDIA's Physical AI, improving adaptability in dynamic industrial environments.
NVIDIA's open platform for AI training, simulation, and in-vehicle computing is enabling the world's robotaxi leaders to scale autonomous vehicle fleets, driving the commercial breakthrough of physical AI.
Someone gave GPT-6 Astra a robot, paintbrush, and camera, and it taught itself to draw the Golden Gate Bridge, improving with each attempt in a stunning timelapse.
Humanoid robots are showcased prominently at IFA Berlin, with consumer electronics companies now integrating them into major tech exhibitions alongside traditional devices like TVs and phones.
Antioch Robotics raises $32 million in Series A funding to build simulation infrastructure that enables Physical AI teams to test releases like software teams, integrating simulation into CI/CD for reproducible and scalable testing.
Actor Labs and Physical Intelligence are hosting act-athon, a physical AI hackathon in Mountain View, CA, where teams will deploy models like pi 0.7 onto robotic embodiments over 36 hours.
An opinion piece arguing that reproducibility in machine learning research is becoming a lost cause due to the rise of physical AI requiring expensive hardware, unverifiable performance claims from big tech companies, and competitive incentives that discourage authors from sharing code.