Robots have always found it challenging to work with humans and vice versa. Two people on the cutting edge of enhancing that relationship came to us for TC Sessions: Robotics to talk about the present and future of human-robot interaction: Veo Robotics co-founder Clara Vu and Robust.ai founder Rod Brooks (formerly of iRobot and Rethink robotics).
Part of the HRI challenge is that while we already have robotic systems that are very capable, the worlds they operate in are still very narrowly defined. Clara said that as we move from “automation to autonomy” (a phrase she emphasized that she didn’t invent), we are adding both possibilities and new levels of complexity.
“We’re going … from robotic systems that do exactly what they’re supposed to do or can sense a very specific, very low level, to systems that have a little more autonomy and understanding,” she said. “The system my company is building wouldn’t have been possible five years ago because the sensors we use and the processors we use to crack that data simply didn’t exist. So because we have better sensors and more processing capabilities, we’re able, as you said, to understand a little bit more about the world we’re in and take the level of robot performance to the next level.”
Brooks highlighted the under-the-hood complexity in the “no-code” tools his new company places in warehouses.
“We have a lot of code; the customers don’t have to code – that’s the difference,” he said. “You know, 80% of all warehouses in the US have no automation, while a conveyor belt would count as automation. 80% don’t even have that. We’re trying to put robots, intelligent robots, in, we don’t want to ask them to understand intelligent robots and programming and things like that when they haven’t had automation. So we have to make it easy for them.”
It’s part of a shift in the overall ecosystem that Brooks sees happening, which has to do with the steady advance of computational improvement giving way to a more creative era.
“I have said that we are in a golden age of computer architecture. Because since 1965, everyone had to adhere to Moore’s Law. They knew they were doubling the speed, doubling the memory, doubling this on this day, or their competitors would get them. So they couldn’t do anything new and weird,” he explained. “With the end of Moore’s Law, they now have to do new and weird things. These are things we couldn’t do two years ago. And that’s because there’s a change in computer architecture.”
That might be a good thing, because the things expected of robots are also getting weirder, relying more and more on an AI that isn’t quite up to the task.
“I think in robotics in general, the robotics problems get exponentially more difficult as the environment is more uncontrolled and the task is more diverse,” said Vu. “So something that would be very simple in a single task and a fixed environment becomes AI complete, shall we call it, in an outdoor environment that is unstructured. And it’s not just a little harder. It’s not just, you have this today and in a few years you will have that. It could be decades more difficult.”
Turning to the field of collaborative robots, or cobots, Brooks remembered his time at Rethink Robotics as valuable and even successful, despite the company eventually going bankrupt.
(A side note for his actual answer: “First off, I have to say Clara is smarter than me because I tried to get her to work, she was a consultant at Rethink, but she didn’t want to come. So she’s smarter than me. . Where were we?”)
“I call Rethink a complete artistic success,” he said. “It changed what people thought was possible and other people do. We were in a way too early and we made a fatal mistake in not sticking to the original idea, which was to put robots not in places where robots already were, but to put them in other places. Because once we went where they already were, there were expectations of what to do. And that pulled us away from what our primary mission was.”
Vu agreed, saying that Rethink had shocked the industry even though it wasn’t a commercial success, noting that the idea for Veo and its co-founder both essentially stemmed from Brooks’ company:
“The idea of collaborative robotics, as far as I know, came from Rethink. How can robots be different from what they are? What could they do that they can’t do today? And most importantly, how could robots work with humans? And how could that actually make the robots more valuable?”
Veo’s goal is to take the cobot idea to the next level:
“Cobots have completely changed the industry. I think there’s 200,000 of them, it’s growing by 30% a year — all the major robot manufacturers are now making cobots too,” she said. “And we’re really trying to take the next step and say, you know, what the ideas behind Rethink have done for smaller, lighter robots… We want to do that for the big powerful robots too, and the way that’s by computer vision, which was not possible 10 years ago.”
We’ve covered many more topics in our discussion, so be sure to check out the full interview below.