Designfax weekly eMagazine

Subscribe Today!

Archives

View Archives

Partners

Manufacturing Center
Product Spotlight

Modern Applications News
Metalworking Ideas For
Today's Job Shops

Tooling and Production
Strategies for large
metalworking plants

Toyota says it has made huge breakthrough in teaching robots new behaviors

Robotic scientists at Toyota's Research Institute and its partners have developed new techniques to teach dual-arm robots advanced behaviors quickly with predictable repeatability.

 

 

Toyota's Research Institute (TRI) recently announced a breakthrough generative artificial intelligence (AI) approach based on Diffusion Policy to quickly and confidently teach robots new, dexterous skills. This advancement significantly improves robot utility and is a step toward building Large Behavior Models (LBMs) for robots, analogous to the Large Language Models (LLMs) that have recently revolutionized conversational AI.

Previous state-of-the-art techniques to teach robots new behaviors were often slow, inconsistent, inefficient, and limited to narrowly defined tasks performed in highly constrained environments. Roboticists needed to spend many hours writing sophisticated code and/or using numerous trial-and-error cycles to program behaviors.

TRI has already taught robots more than 60 difficult, dexterous skills using the new approach, including pouring liquids, using tools, and manipulating deformable objects. These achievements were realized without writing a single line of new code; the only change was supplying the robot with new data. Building on this success, TRI has set an ambitious target of teaching their robots hundreds of new skills by the end of the year and 1,000 by the end of 2024.

"This new teaching technique is both very efficient and produces very high-performing behaviors, enabling robots to much more effectively amplify people in many ways," said Gill Pratt, CEO of TRI and chief scientist for Toyota Motor Corporation.

Part of the excitement at TRI is that robots can be taught to function in new scenarios and perform a wide range of behaviors. These skills are not limited to just "pick and place" or simply picking up objects and putting them down in new locations. TRI's robots can now interact with the world in varied and rich ways, which will one day allow robots to support people in everyday situations and unpredictable, ever-changing environments.

"The tasks that I'm watching these robots perform are simply amazing," said Dr. Russ Tedrake, vice president of Robotics Research at TRI. "Even one year ago, I would not have predicted that we were close to this level of diverse dexterity."

Tedrake, who is also the Toyota Professor of Electrical Engineering and Computer Science, Aeronautics and Astronautics, and Mechanical Engineering at MIT, explained, "What is so exciting about this new approach is the rate and reliability with which we can add new skills. Because these skills work directly from camera images and tactile sensing, using only learned representations, they are able to perform well even on tasks that involve deformable objects, cloth, and liquids -- all of which have traditionally been extremely difficult for robots."

How the new robot learning is accomplished
TRI's robot behavior model learns from haptic demonstrations from a teacher, combined with a language description of the goal. It then uses an AI-based Diffusion Policy to learn the demonstrated skill. This process allows a new behavior to be deployed autonomously from dozens of demonstrations. Not only does this approach produce consistent, repeatable, and performant results, but it does so with tremendous speed.

Key achievements of TRI's research for this novel development include:

• Diffusion Policy: TRI and its collaborators in Professor Shuran Song's group at Columbia University developed a new, powerful generative-AI approach to behavior learning. This approach, called Diffusion Policy, enables easy and rapid behavior teaching from demonstration. Song is assistant professor of Electrical Engineering, by courtesy, of Computer Science at Stanford University and adjunct associate professor of Computer Science at Columbia University.
• Customized Robot Platform: TRI's robot platform is custom built for dexterous dual-arm manipulation tasks with a special focus on enabling haptic feedback and tactile sensing.
• Pipeline: TRI robots have learned 60 dexterous skills already, with a target of hundreds by the end of the year and 1,000 by the end of 2024.
• Drake: Part of TRI's (not so) secret sauce is Drake, a model-based design for robotics that provides the organization and its collaborators with a cutting-edge toolbox and simulation platform. Drake's high degree of realism allows them to develop in both simulation and in reality at a dramatically increased scale and velocity than would otherwise be possible. Their internal robot stack is built using Drake's optimization and systems frameworks, and they have made Drake open source to catalyze work across the entire robotics community.
• Safety: Safety is core to the robotics efforts at TRI. They have designed their system with strong safeguards, powered by Drake and their custom robot control stack, to ensure every robot respects safety guarantees such as not colliding with itself or its environment.

Diffusion Policy has been published at the 2023 Robotics Science and Systems conference. Additional technical information can be found on TRI's Medium blog.

Source: Toyota Research Institute

Published November 2023

Rate this article