The Hallway Track
12 tracked signals on embodied-ai.
NVIDIA's ENPIRE framework lets coding agents run a closed-loop self-improvement process for real-world robots, hitting 99% on dexterous tasks.
“Frontier coding agents can autonomously develop a policy to achieve a 99% success rate on challenging, dexterous manipulation tasks in the real world, such as PushT, organizing pins into a pin box, and using a cutter to cut a zip tie,”
Perceptron AI wants to unify VLMs, VLAs, and world models into 'embodied foundational models' for real-time physical-world interaction.
“we want to move away from the distinction between VLM, VLA, world models, whatever you want to call it, to what we call embodied fundamental models”
Physical Intelligence robots now perform complex household tasks in previously unseen environments
“we showed how robots can do useful tasks in environments in rooms they've never been in before”
NVIDIA Cosmos-H-Dreams enables real-time generative simulation for surgical robotics training.
Google DeepMind and Boston Dynamics are partnering to bring frontier AI breakthroughs into physical robotics
“you have this vision language and action model and you could like talk to the robot and you can ask it to do things and you get all this like understanding for free in robotics”
Google DeepMind launches Gemini Robotics ER 1.6 for spatial reasoning in physical robots
“We aren't just teaching machines to see the world, but we're enabling them to perceive it with advanced spatial awareness and understanding.”
MoonLake AI is building modeling infrastructure for practical physical, embodied AI grounded in causality.
“The North Star for AI, and indeed for cognitive science, will always be to understand and develop how to create embodied intelligence.”
Hugging Face shows deploying models from the Hub to robot hardware using Strands Agents and LeRobot.
Robotics remains unsolved in 2026 despite a decade of 'next year' predictions and new breakthroughs.
“it's definitely the year of the demos. I know that for certain.”
NVIDIA combines imitation learning and goal-based RL to train parkour AI on just 30 seconds of data
“Systems that copy us are beautiful, but brittle, repetitive. Systems that chase goals adapt, but often stop moving like us.”
Hugging Face releases Grabette, an open system for recording robot manipulation data
OpenMind's OM1 is an open-source, hardware-agnostic orchestration layer bringing LLM cognition to robots.
“we are building software that brings cognition to the robots”