DigestAI news desk

Cut through the AI noise.

Robotics & Physical AI3 min read

RAI robot demonstrates whole-body manipulation in complex tasks

Researchers at ETH Zurich’s Robotic Systems Lab showcased a humanoid robot named RAI performing whole-body manipulation in a video featured by IEEE Spectrum. The robot uses a contact-centric learning framework combining motion imitation and reinforcement learning, trained on human motion data to adapt to new environments. It can handle objects like tires but may also shift focus to other items…

1 source

Key points

  • ETH Zurich’s RAI robot uses contact-centric learning to manipulate objects with whole-body movement
  • System leverages human motion data for adaptation in unstructured environments, per lab
  • Video Friday series highlights robotics demos, including industrial and drone-focused innovations

The demo highlights progress in humanoid robotics, though the lab notes ambiguity in whether the robot actively engages its torso—an ability humans rely on for heavy or awkward objects. The video is part of IEEE Spectrum’s Video Friday series, which also spotlights other robotics innovations, including a vision computer for drones and industrial climbing robots.

The story so far

2 episodes →
  1. RAI robot demonstrates whole-body manipulation in complex tasksthis story
Full story from IEEE Spectrum Robotics · by Evan AckermanOpen source ↗

Video Friday: Spot Plays to Show Its Whole Body Manipulation

IEEE Spectrum Robotics · 30 October 2026

Loading the full article…

This text was published by IEEE Spectrum Robotics and written by Evan Ackerman. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗

Topics · follow one to build your own front page

The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.

Comments

via GitHub Discussions

More in Robotics & Physical AI

All →

Related stories