The Army wants to fight with AI-driven ground systems. Where does development stand?
Oct 29, 2025
Overland AI CEO Byron Boots discusses the challenges of developing autonomous ground systems for the Army at AUSA 2025.
View Video Transcript
0:00
Welcome back. Autonomy is an increasingly popular theme among companies developing military systems, including at Overland AI, which is creating autonomous ground systems for the Army
0:10
To talk about the process, we sat down with Overland CEO Byron Boots. Here's part of the conversation
0:17
What is Overland AI? Talk a little bit about your company and what it is that you've been working on
0:23
Sure. Thanks for having me, Jen. So Overland AI is ground autonomy for defense. We build
0:29
autonomy software, so a software stack which can take an uncrewed vehicle and turn it into an
0:35
autonomous asset. And we also build autonomous vehicles themselves, so a fully vertically
0:40
integrated solution, which is the ground vehicle, the autonomy stack, and it can carry payloads
0:46
and move them across the battlefield. You know, so I think it was maybe a year ago
0:51
I had a conversation with General Glendine from PEO Ground Combat Systems
0:59
And, you know, at the time they were assessing autonomy for ground robotics, ground vehicles
1:07
And he seemed to say, you know, it's not really ready for us yet
1:13
How have things evolved? Would you say that the attitude might be a little bit different a year later
1:19
What were some of the hurdles that you were encountering that made it that much harder to apply the autonomy to what the Army needs it for versus a taxi, if you will
1:32
So just in the last year and a half or so, there's been a lot of really exciting advancements in ground autonomy and us really just showing people what's possible
1:43
So the whole reason that the DARPA program was stood up was because of the challenges
1:48
of trying to move vehicles through difficult off terrain And some of those challenges are having to actually contend with really complex ground geometry like ditches and berms and things like that
1:59
Kind of recognize what that is. Yeah, yeah, exactly. Like, can you see it? Can you model it
2:03
And then how does a vehicle move through it? And then also things like dense vegetation, which occupies space, so it's kind of like
2:10
an obstacle, but you can also push through it, so it's not really. So how do you deal with things like that
2:15
And these were difficult research questions a few years ago and something that a lot of autonomous systems had a lot of trouble contending with
2:22
So through the course of working with DARPA, we really developed a solution which was able to handle a lot of these sorts of things
2:29
Then coming out of DARPA and forming the company, our challenge then was actually showing the broader Army and Marine Corps what we were capable of and actually just getting eyes on it
2:39
And so if you look at the last year and some of the comments that Army leadership was making
2:45
I think they hadn't seen what was actually possible at that point. And so what we did is we came into a program called Ground Vehicle Autonomous Pathways
2:54
We entered into that program about halfway through the course of it and immediately started to just demonstrate for the Army what was possible
3:03
So last year there was a lot of talk about teleoperation and essentially soldiers remote controlling ground vehicles
3:10
because the autonomy wasn't really where it needed to be. But I think what we've been able to do
3:16
is show the real power of this kind of new type of ground autonomy
3:22
what it's capable of, be able to show that in the field. And now I think what you're seeing
3:26
is a lot of people getting really excited about autonomy, like it actually works, it exists
3:30
And so you're not having those conversations anymore. It's much more about how do we take that autonomy
3:35
move it onto different platforms, how do we coordinate it with payloads
3:39
and actually start to use it to provide more capability to our warfighters
#news


