Over the next few weeks, TechCrunch’s robotics newsletter Actuator will feature Q&A with some of the top minds in robotics. Subscribe here for future updates.
part 1: Matthew Johnson Robertson, Carnegie Mellon University
This week, we have two people. Russ Tedrake and Max Bajracharya of Toyota Research Institute shared this work. Tedrake is Vice President of Robotics Research at TRI. He is also the Toyota Professor of Engineering and Aeronautics/Astronomy in the MIT Department of Electrical Engineering and Computer Science, and the Department of Mechanical Engineering. Bajracharya is Senior Vice President of Robotics at TRI. He previously served as director of robotics at the institute.
What role will generative artificial intelligence play in the future of robotics?
Russ Tedrake: Generative artificial intelligence has the potential to bring revolutionary new capabilities to robotics. Not only are we able to communicate with robots in natural language, but being connected to internet-scale language and image data gives robots a more powerful understanding and reasoning about the world. But we’re still in the early stages; more work needs to be done to understand how to incorporate imagery and language knowledge into the type of physical intelligence needed to make robots truly useful.
What are your thoughts on the humanoid form?
Max Bajrajacharya: Places where robots can assist humans are often designed for humans, so these robots may need to adapt to and work in those environments. However, that doesn’t mean they need to be humanoid (two arms, five-fingered hands, two legs). Simply put, they need to be compact, safe, and capable of performing human-like tasks.
After manufacturing and warehouses, what is the next major category of robotics?
Max Bajrajacharya: I saw huge potential and need in agriculture, but the outdoor and unstructured nature of many tasks was very challenging. Toyota Ventures has invested in several companies, including Burro and Agtonomy, that are making good progress in bringing autonomy to some people for initial agricultural applications.
How far are we from truly universal robots?
Russ Tedrake: I’m optimistic that the field can steadily advance from the relatively niche robots we have today to more general-purpose robots. It’s unclear how long this will take, but flexible automation, high-hybrid manufacturing, agricultural robots, point service robots and possible new industries we haven’t yet imagined will benefit from increasing levels of autonomy and increasingly versatile capabilities.
Will home robots (beyond vacuum cleaners) boom in the next decade?
Max Bajrajacharya: 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 evolving rapidly.
What important robotics stories/trends aren’t getting enough coverage?
Russ Tedrake: We hear a lot these days about generative artificial intelligence and the incredible advances and investments in hardware. However, many of these successes have been achieved through the quiet revolution we see in simulations. Just a few years ago, most robotics experts would have said it was impossible to train or test computer vision systems in simulation; now this is standard practice. Some researchers are still skeptical that we can develop a control system, such as a dexterous hand, entirely in simulation and have it function in reality, but the trend is increasingly moving in that direction. Huge investments from companies like Nvidia, Google DeepMind and TRI are helping make this happen.