Robotics Q&A with Max Bajracharya and Russ Tedrake of the Toyota Research Institute | TechCrunch

Over the coming weeks, TechCrunch’s robotics newsletter, Actuator, will host Q&As with some of the greatest minds in robotics. Subscribe here for future updates.

Part 1: CMU’s Matthew Johnson-Roberson

This week we have a double. Russ Tedrake and Max Bajracharya of the Toyota Research Institute share the work. Tedrake is TRI’s vice president of robotics research. He is also the Toyota Professor in the Department of Electrical and Computer Engineering, Mechanical Engineering, and Aero/Astronomy at MIT. Bajracharya is TRI’s senior vice president of robotics. He was previously director of robotics at the Institute.

What role(s) will generative AI play in the future of robotics?

Russ Tedrake: Generative AI has the potential to bring revolutionary new capabilities to robotics. Not only are we able to communicate with robots in natural language, but connecting to Internet-scale linguistic and image data gives robots much more robust understanding and reasoning about the world. But we are still only in the early days; more work is needed to understand how to ground knowledge of images and language into the kinds of physical intelligence needed to make robots truly useful.

What do you think of the humanoid form factor?

Max Bajracharya: Places where robots can help people tend to be designed for people, so these robots will likely need to adapt and work in those environments. However, this does not mean that they have to take a humanoid form (two arms, five-fingered hands, two legs and a head); they simply need to be compact, safe, and capable of performing human-like tasks.

After manufacturing and warehouses, what is the next big category in robotics?

Max Bajracharya: I see a lot of potential and need in agriculture, but the outdoor and unstructured nature of many tasks poses a real challenge. Toyota Ventures has invested in a few companies like Burro and Agtonomy, which are making good progress in bringing autonomy to some early-stage agricultural applications.

How far away are true multi-purpose robots?

Russ Tedrake: I am optimistic that the field can progress steadily from the relatively specialized robots we have today to more versatile robots. It’s unclear exactly how long it will take, but flexible automation, high-mix manufacturing, agricultural robots, point-of-care robots, and likely new industries we haven’t yet imagined will benefit at increasing levels. of autonomy and increasingly general capacities. .

Will domestic robots (beyond vacuum cleaners) take off in the next decade?

Max Bajracharya: Homes remain a difficult challenge for robots because they are so diverse and unstructured, and consumers are price sensitive. The future is difficult to predict, but the field of robotics is advancing very quickly.

What important story/trend in robotics isn’t getting enough coverage?

Russ Tedrake: We hear a lot these days about generative AI and the incredible advancements and investments in hardware. However, many of these successes have been made possible by the quiet revolution we have witnessed in simulation. Only a few years ago, most roboticists would have said that it was impossible to train or test a computer vision system in simulation; now it is common practice. Some researchers are still skeptical about the possibility of developing a control system, for example for a dexterous hand, entirely in simulation and making it work in reality, but the trend is increasingly moving in this direction. Big investments from companies like Nvidia, Google DeepMind and TRI are helping to make this happen.