Robots do backflips and somersaults, but can they actually do work? Our very own Julia Borsten is at the Aspen Ideas Festival with Agility Robotics CEO Peggy Johnson. Julia. Thanks so much, Tyler. And I'm joined now by Peggy Johnson, the new CEO of Agility Robotics and your first interview since taking over as CEO in March, previously CEO of Magic Leap. And today you are announcing some big news on our air. Your Digit robot, which is a humanoid robot, is in a new partnership with GXO Logistics. Tell us this news. Yeah, this is big news. It's an industry. 1st Digit is stepping in and picking up tasks at GXO. They're a third party logistics. Operator just outside of Atlanta will be moving product within their warehouse. This is kind of the dull, repetitive sorts of tasks that's hard to find people to step in and do. But we are on the clock now. So these these robots are are doing the work that would be done by workers previously. Tell us a little bit about agility. Your valuations over a billion dollars. You've raised over $100 million. What makes your humanoid robots so different not just from the other robots? That are being deployed in warehouses, but other humanoid robots that we've seen demos of, right. So we're the first ones being put to work. We are stepping in and operating in those spaces that humans previously operated in were different in a couple ways. One is that we have actual data from environments that the inside the customer facility turns out that's very valuable data to help train AI. Models that will then teach new skills to digits. So that'll help open up new markets going. So you're talking about teaching digit. This is obviously AI powered, but how autonomous and independent are these robots and what do you see them being able to do as AI technologies continue to advance so quickly? They are Digit is fully autonomous. There's no cord behind it. It is still inside of a work cell. So it works in a work cell for safety you don't want. BI, at this point in time, controlling the robot. So we use traditional methods to control the robot, but we can layer in AI on top of that to help teach it new skills along the way. So here at Aspen Ideas, there's a lot of talk about AI, but there's also a lot of talk about regulation. You talk about you don't want AI controlling the robot. We've heard people warn about robots taking over, our autonomous robots taking over sort of the worst case scenario of the future of AI. What do you think about regulation? Should AI be regulated? And what would regulation mean to your AI powered robots? It absolutely should be regulated, particularly in the humanoid space. These are, you know, human sized, very large devices. And you wanna make sure that the human is always top of mind. And so we there's a moment in time now as AI and robots merge for governments to work on the proper regulation to make sure that everybody follows the guidelines. As these devices are starting to be introduced into our environments. And so you think this is gonna happen at A at a federal level, national level? Do you think it's gonna happen in international level? There's already national standards for certain types of robots. Our team is helping and has a voice in the standards for humanoids going forward. So that standard still is being formulated, but now is the time to get it right. So let's talk about scaling your business. Right now. You're making this big partnership and deploying these robots, putting them to work in this logistics. Company, you mentioned you have a partnership with Amazon, but what's the future? What other industries, types of companies could use your robots maybe to replace workers or certainly to accelerate the pace of their growth? Well, there's a lot of tasks, I would say tasks not the full job that are just repetitive, dull, dirty, sometimes backbreaking, a lot of injury prone tasks that humanoids can step in and take off the plate of the human and that frees the human up to learn. Whose skills to actually be the manager of the robots, they can step into a digital job based on on that. And so it's it's really it's going to change the face of labour going forward, but I wouldn't think of it as taking the job, but as augmenting as humans job.
Business Development Executive l Zoom Robotics I Former Director @ A Mark Cuban Company I Inventor I Master's in Engineering Management I Georgetown University
Meet Figure, a leader in the race to build commercially viable humanoid robots.
For episode 20 of S³, we got the world’s first close up look at what they’re building.
We strongly believe in the commercialization potential of humanoid robots. Completing menial labor tasks initially and then growing skill sets to evolve to more complex actions. To get there, it will be challenging and slow, but look at how far humanity has come over the past century. I'm sure those on horseback never imagined the first motor car, let alone that an automobile would one day handle 0-60 mph in less than 3 seconds.
Figure's view that "labor is a choice" certainly rings true. If humanoids handle basic labor tasks, imagine the free time you'd have to learn new skills and advance technologies. It could help to level up humanity in its entirety if we saved our time from performing that basic labor.
As Brett stated, "the only way out is through", so let's "get out and start building". Let CHASSY help support you to move fast and scale quickly so that we can all realize that future together. One where having a humanoid is just like having a smartphone. Reach out to Chuka Okoye for more details or request a demo (https://lnkd.in/g7TjqWXJ).
#roboticsplatform#humanoids#humanoidfleet#figure#savemytime
Meet Figure, a leader in the race to build commercially viable humanoid robots.
For episode 20 of S³, we got the world’s first close up look at what they’re building.
Love Brad Porter’s thoughts on robots that distill a lot of powerful insights into minimal space. My own futurist predictions, based on very little recent work with robots, a lot of “transfer learning”: 1. ML can chew up the robot compute problem and eat it up for breakfast. If we can train neural networks to fold proteins, it will be able to handle however many degrees of freedom. 2. To me, sensing and actuation are the laggard technical ingredients, and so current solutions will be preoccupied with delivering useful functionality within these constraints (much like Doom/Quake had to make simplifications to make a common PC capable of rendering a 3D game). This means we are going to see more success in specialized robots for a while. 3. When sensors/actuators are no longer the bottleneck, then we will continue to have specialized robots just like we have specialized compute (airplanes are after all a bit faster than other kinds of machines at flying!), but “general purpose robots” will make more and more sense. They will be hugely wasteful from a robot design perspective in the same way that using an LLM answer a question is wasteful, or having an out-of-order superscalar execute an addition instruction instead of slapping down an adder in VLSI. But they will allow you to solve a larger class of problems with one device.
CEO & Founder Collaborative Robotics. AI & robotics leader. Formerly Distinguished Engineer at Amazon and CTO at Scale AI.
I took some time to write up my thoughts on the humanoid robot space. At Cobot, we see a future with more and more capable robots. But we have some conviction those won't look humanoid. Click through to read the details. Feel free to debate in the comments, I'll respond.
Here are the top 10 new humanoid robots. When considering all the tasks they can do and the pace of development, the robots are going to snatch way more jobs than AI ever will.
Flexible (Warehouse) robots will replace Manpower in the warehouses hopefully quite soon. They are faster, more accurate, without performance changes. The communication and motivation challenges disappear, but the working resource displacement to balance the flows will stay. There will be needed only the Flexible Resource Management Concept (and the appropriate AI training of course) to manage robots in order to balance (material) flows in the different types of warehouses.
Chief of Robotics Strategy | MSME
3wCongratulations. Good information.