Topic
Humanoids, robot foundation models, manipulation, autonomy, and embodied AI.
Latest Signals
July 2026 saw significant developments in robotics, including large funding rounds, novel robot training methods, and new robot models. These stories captured strong interest from the robotics community.
Five physical AI infrastructure platforms are identified as key to addressing challenges and forming the foundation of the emerging physical AI tech stack in robotics. These platforms are highlighted by Lightwheel and discussed by The Robot Report.
Google has introduced Gemini Robotics 2.0, featuring three models with improvements in dexterity and safety. Currently, only one model is publicly available.
Teradyne Robotics reported a 33% year-over-year revenue increase in Q2 2026, with significant growth attributed to the U.S. market and AI-related demand.
Google DeepMind's latest Gemini Robotics AI model can control whole-body motions of humanoid robots, extending from feet to fingertips. This is an advancement from the previous model, which only managed upper body movements.
The US government's ban on foreign-made robots aims to protect domestic robotics but may ultimately hinder the US robotics sector. The policy could disrupt supply chains and slow innovation rather than boost local development.
The US government has banned new imports of foreign-made humanoid robots, robot dogs, and solar inverters, primarily targeting Chinese products. This move is driven by concerns over national security risks associated with these technologies.
GMEX Robotics has acquired shares in MediaMeta.ai to integrate social intelligence capabilities into its service robots. This move aims to enhance the robots' interaction and responsiveness in social environments.
A school district in upstate New York has paused plans to introduce an AI-powered humanoid robot teacher following concerns from state officials, teachers, and residents. The decision reflects apprehension about integrating such technology in classrooms.
Hacker News surfaced this AI signal from theguardian.com: FCC ban humanoid robots from China.
Researchers explore transferring a text-conditioned grasp detection model to speech inputs for humanoid robots using ALBEF, aiming for data-efficient multi-modal understanding. This approach addresses the gap between vision-language models and natural speech interaction.
RoboBusiness 2026 will be held in October in Santa Clara, marking 20 years of robotics industry events. Registration is now open for attendees.
Two Scythe co-founders have launched Agency Tool Company to develop tools that make over-the-air (OTA) software updates easier for robotics developers. Their experience in commercial robotics informs this new venture.
NVIDIA's Cosmos-H-Dreams platform introduces real-time generative simulation capabilities tailored for surgical robotics. This advancement aims to enhance the precision and adaptability of robotic surgical systems through improved simulation techniques.
Researchers developed a lightweight episodic memory system combining vector-based semantic retrieval with an LLM dialog controller for the humanoid robot head Kim. This approach addresses the lack of memory across sessions in social robots, supporting more natural, persistent relationships.
NVIDIA showcased breakthroughs in AI and graphics that enhance real-time simulation and robotics capabilities. These advancements include open models that improve media, content creation, and robotic systems.
Agility is opening a new training center for its Digit robots in Fremont, California.
There are many words that I would never, ever use t...
NVIDIA introduced the T3000 and T2000 modules based on the Thor architecture to support compact, power-efficient AI supercomputers for mainstream robotics and edge AI. These modules aim to enable real-world deployment of general-purpose robots and autonomous machines.
Home to leading manufacturers, robotics pioneers, infrastructure builders and iconic gaming companies, of course, Japan is one of the world’s centers of AI — building across the full stack with NVIDIA technologies. This week NVIDIA and its partners in Japan a...
The article discusses how large language models have provided a working recipe for artificial intelligence, emphasizing the importance of pretraining. It explores the development of a foundational software stack to enable general-purpose robots to perform diverse tasks.
General Intuition is betting millions of hours of video game data can train the foundation models for physical AI, making it easier to build smarter robots with minimal real-world data.
Agility Robotics is going public through a SPAC while focusing on execution rather than chasing high valuations. The CEO cautions that humanoid robots for home use are not imminent.
The startup, Proception, is taking a unique approach to collecting training data to tackle one of the hardest problems in robotics: hands.
Agility Robotics, the humanoid robotics startup that spun out of Oregon State University in 2015, expects to generate $620 million in proceeds.
Hacker News surfaced this AI signal from avikde.github.io: Show HN: A reasoning hierarchical robotics pipeline you can run in the browser.
Humanoids, robot foundation models, manipulation, autonomy, and embodied AI.