The Robot Report Podcast
00:21:22 · Mike Oitzman, Aaron Nathan
[00:21:22] S1: Yeah. Because I follow so many mobile robots, especially outdoor robots now in in ag, for example. You know, we're talking about one of the most important and often overlooked challenges with outdoor robotics, and that's knowing where you are. You know, Aaron has spent the last two decades, as we're gonna hear, working on robotics and autonomous vehicle systems. You know, he started his career at Cornell's DARPA Grand Challenge team, and later he founded Point One to build reliable high precision location technology. And in this conversation, Aaron's going to explain to us why location is foundational to robotics and how RTK can improve positioning from meters to centimeters and why ArcTK alone isn't enough in environments like orchards, urban canyons, and construction sites to to to keep you located. He's also gonna share with us how it can scale from robotics pilot to a production fleet. So again, if you're if you're building mobile robots, it's definitely gonna be an interesting conversation. So let's listen to that now. I wanna welcome Aaron Nathan to the show today. He's the CEO and cofounder of Point One Nav. Aaron, welcome to the show. [00:22:37] S2: Thank you for having me. It's great to be here.
00:22:37 · Aaron Nathan, Mike Oitzman
[00:22:37] S2: Thank you for having me. It's great to be here. [00:22:39] S1: Yeah. We're excited to dig under the hood of navigation technology and and it's, you know, it's one of those things that we recognize is so important for mobile robots, especially outdoor mobile robots. I think that's the first place that you and I sort of stumbled onto each other was talking about agriculture robots. You've been you've been enabling ag robots to be more precise. We'll come back and talk about that in a minute. So you've been building, you know, location for robots and autonomous vehicles for twenty years now. You started Cornell's DARPA Grand Challenge team in 2005. I'm sure that was an amazing process. And then you started Point One Nav in 2016. So tell us, you know, what have you learned about what's hard for a machine to know where it is in the world today? Let's start there.
00:23:26 · Aaron Nathan
[00:23:26] S2: Yeah. For sure. And, again, thanks for having me here, and it's been great to kind of be on the journey together, so to speak. And, yeah, I mean, 2005 was a long time ago. And, you know, the thing is that the robotics problems back then, they're still very much the same robotics problems we have today because we are still in the same physical world the robots have to work in. It turns out that location is one of those things that you really, really have to get right, early in the design process of, like, your robotic system. Otherwise, everything else downstream, kinda falls apart. And so back in 2005 when we were doing the DARPA grand challenge, this was, a highly ambitious, thing. At the time, I was a sophomore, I think, in university, and, you know, we had this brilliant idea of like, oh, yeah. No problem. We can make a full sized car drive itself across the desert for over a 150 miles, which was more or less the challenge. And, you know, you tell a bunch of college kids, like, yeah, you get a million bucks if you can pull it off. Like, we're off to the races. And what we learned very quickly was that location was really the key to that problem because driving this one was off road, there so wasn't other traffic or other vehicles to avoid. It was really a problem of, like, how do you stay on the road? And the challenge was not necessarily, like, how do I get to some level of accuracy? Like, it's actually more how do you do it all the time? Because if you think about, like, driving for an endurance, it's about making the right decision, like, all the time. Not 99% of the time, but, like, all the time. And that really highlighted to me, like, the importance of location technology and how it has to work, like, all the time. And, you know, fast forward now twenty years, which is wild, what we've seen is that this, has only become more true. Basically, like, you know, the vehicles or other robotic systems, outdoor robots, agro robots, they're, developing more and more capabilities, like, you know, whether they're able to now, you know, very precisely... We have some customers who are literally, like, burning weeds with lasers or other, you know, ultraviolet lights, all kinds of crazy things in agriculture, similar things in construction and mining, you know, guys delivering packages with drones to, you know, your doorstep.
00:26:04 · Aaron Nathan, Mike Oitzman
All of those things are enabled because now we have a lot more capability in the, like, general robotics world, which actually pushes further on this problem that, like, location has to work all the time because you have robots trying to do more things, and so the stakes are higher. The, the bounds of, like, hey. Can you make a mistake, are actually smaller. So if anything, like, the problem has become even more important. And I think, you know, that's that's something that the industry is starting to really pick up on is, hey. This is something that's not, like, a simple thing on the side. When I actually get to scale, it becomes, like, a foundational block that, like, really has to work all the time. Mhmm. [00:26:53] S1: We're gonna dive deeper into your technology a little bit later on in in the conversation here. But across that... The scope of the last twenty years, I mean, I think of the evolution what we have as consumers. Our cell phones can now tell us where we are. Google Maps tells us... Gets us from point a to point b today. But along the way, it seems like GPS has been ubiquitous along that whole path. But you know, what's really changed in the last twenty years with GPS or... And or maybe some of the other technologies that that are helping support this?
00:27:28 · Aaron Nathan
[00:27:28] S2: Yeah. I mean, satellite navigation is like a broad topic, which is GPS is like the US government's flavor of that. It has been available, you know, kinda since the nineties. There is this whole thing where the government, there used to be, like, a scrambling that was on top of GPS that they removed and signed into law that they will not use that scrambling. And then, you know, the whole industry of... I'm sure everyone remembers, you know, putting that sticky Garmin thing on your windshield to navigate around here. That was that was like a whole whole thing, I think, when I was in high school or something. So... But, yeah, so satellite navigation has been evolving but slowly. And, I mean, part of this is that GPS satellites are very complex, very expensive, and they technically are a DoD asset first. You know? So there's constraints that they have on things that they need to be able to do all the time. But that said, you know, we moved from a world where GPS and then the Russian system called GLONASS were really the two major systems that were out there. Then the European Union, they launched Galileo, which was, you know, basically a full navigation system similar to GPS. And then, you also have the Chinese that launched Beidou, which is a very advanced navigation system that also works similarly to GPS. And so, your phone, when you're walking around, uses all of those. It's not just using GPS. It's actually using all of these different, what we call, Geonetscess systems. And the accuracy you get from them is higher just because you have more satellites, more points. The other big thing that's happened is GPS and all of its cousins, there's... Traditionally was one frequency that they would use. And without getting too much into the weeds, like, having two frequencies or three frequencies now, allows them to be actually much more precise because of how the physics works out. And so, this is a big thing now that, like, literally consumer level devices like what's in your phone are able to use, not just one, but two frequencies. And, you know, consumers have seen this, like, Apple actually made, like, a big deal about it when they launched the latest Apple Watch, which was like, yeah. What do you see?
00:29:56 · Aaron Nathan, Eugene Demetri
It's like, well, while I'm running, you know, in the city or something, my actual Strava track or whatever it is I'm recording on, like, looks really good because you're actually now able to use those two frequencies. I mean, big technical things that had to happen, like, ultimately do result in these consumer facing improvements. [00:30:18] S3: So, Aaron, let's talk about Point One Nav's upcoming FUSE conference. I understand it's not a product conference, but Randall won about location technology. Why have a conference about location technology?