(Source: TRI Press Room)
The Toyota Research Institute (TRI) is unveiling an electrifying breakthrough in robotics by introducing an innovative training approach that catapults robots from the mundane pick-and-place universe into a realm where they can swiftly grasp and perfect novel, dexterous behaviors. While robots offer promise in assisting with a wide variety of tasks, there's always been a significant barrier in teaching these tasks—until now.
Using TRI's groundbreaking method, robots don't just learn; they observe, absorb, and master complex manipulative tasks overnight, acquiring the agility to wield tools, pour liquids, and even peel vegetables, all without a single change in code or the need for exhaustive programming (Figure 1). Say goodbye to the days of laborious machine learning with millions of training cases and embrace a dawn where robots are not just smart, but astonishingly skilled and impressively adaptive.
Figure 1: “Teaching” the robot cooking skills that it will learn in one afternoon and be able to perform by morning the next day. (Source: TRI)
Russ Tedrake, VP of Robotics Research at TRI, launched a mission in 2016 to propel robots to unprecedented dexterity. Early hurdles led to a game-changing question: "What's the Achilles' heel of our new-age robotics?" The answer? An overreliance on simulation. Enter the Intuitive Physics Project, a daring initiative to unlock human-like abilities beyond the confines of virtual worlds.
Instead of leaning on simulations, the team embraced a radical strategy: imitation learning with a modern spin. Fast-forward to 2022, a bold summer intern introduced a groundbreaking tweak—diffusion models. The models allow efficient learning of reactive policies from a small set of provided training data.
After collecting a set of demonstrations for a particular behavior, the robot learns to perform that behavior autonomously. At the heart of this capability is a generative AI technique called Diffusion Policy1. For example, a teacher demonstrates a small set of skills, and Diffusion Policy generates robot actions based on sensor observations of human movement and natural language. The demonstration runs for several hours after the teacher’s original demonstration. The use of diffusion provides three key benefits:
A sense of touch also allows the robots to learn more easily. TRI Soft-Bubble sensors2 consist of an internal camera observing an inflated deformable outer membrane and go beyond measuring sparse force signals. This allows a robot to perceive spatially dense information concerning contact patterns, geometry, slip, and force. The robots perform well when they can “feel” an interaction with the environment (Table 1).
Table 1: A real-world performance comparison between tactile-enabled and vision-only learned policies. (Source: TRI)
Imagine a world where robots can whisk up your breakfast or flip a pancake effortlessly. That's the horizon TRI is reaching for with its innovative approach. By leveraging a variety of teleoperation tools—from the humble joystick to sophisticated bimanual devices—they're rewriting the rules of robot education. With the magic of cutting-edge generative AI, they've empowered a robot to master an array of intricate skills, like culinary arts and gadget handling. Siyuan Feng, a TRI researcher at the forefront of Diffusion Policy, sheds light on the process: while they guide the robots, they've built resilience into their learning DNA, readying them for real-world hiccups.
The key to receiving dexterous commands is position-position coupling, where a human operator can teach the robot poses and movements with a bimanual input. The robot runs the poses and movements through torque-based Operational Space Control, mimicking the operator’s movements and sending haptic feedback to the operator. And, excitingly, TRI plans to share this dance with the world, hinting at an open-source release.
Simulation, once sidelined for hands-on action, is making a dramatic comeback. TRI is meticulously crafting an immersive curriculum in their virtual "robot kindergarten," both in the tangible world and their digital playground. They've already chalked up over 60 behaviors, and the end goal is robots that can surprise us by pulling off feats they were never explicitly taught.
With their sights set high, TRI is gearing up to unveil hundreds of new robot skills in the near future, aiming for a staggering thousand by the end 2024. And the cherry on top? Their simulation expertise is sowing the seeds for a robot revolution, where a lesson learned by one becomes wisdom for all. The future, it seems, is a dance of metal and AI.
Ben Burchfiel, the head of the TRI dexterous manipulation team, shares, “We're pushing the boundaries with scalable algorithms, drawing inspiration from the leaps in language and image tech.” Yet, the world of robotics is still hungry for a rich and diverse data tapestry, something TRI is passionately crafting. Their vision? To weave large behavior models that blend deep understanding with tactile brilliance, empowering robots to intuitively devise new nimble actions. This journey promises to not only redefine robotics but also ignite fresh innovations in automation, machine learning, and the collaboration between humans and machines.
Sources
Carolyn Mathas is a freelance writer/site editor for United Business Media’s EDN and EE Times, IHS 360, and AspenCore, as well as individual companies. Mathas was Director of Marketing for Securealink and Micrium, Inc., and provided public relations, marketing and writing services to Philips, Altera, Boulder Creek Engineering and Lucent Technologies. She holds an MBA from New York Institute of Technology and a BS in Marketing from University of Phoenix.