It's a group within AI called physical AI. You have a robot that is collecting information around it through various sensors, and then, based on the algorithms it's been programmed with and has learned, it's able to then observe, orient itself, decide what to do and act on it.
In the case of a robot on the moon, it's perhaps helping do the job it's trying to do—unload a spacecraft or create a berm for a landing pad or whatnot—and that's possible because it has that autonomy. Having a human in the loop to try to control that would be almost impossible because of the time delay and all of those things.
That technology becomes highly relevant in other scenarios here on earth, where there might not be GPS, where bandwidth is limited or where there are significant communication delays. The Arctic would be a really good example. For us to deploy autonomous systems to those places to ensure that we can maintain sovereignty, I think, is a really good use case for this kind of technology.