Sending robots to space is the ‘first step’ to bringing AI to everday life: Group Leader and Technologist at NASA JPL and Caltech – Yahoo Finance

Ali Agha, Caltech Project Lead, JPL Nebula Autonomy and AI, NASA Jet Propulsion Laboratory, talks testing Boston Dynamics' Spot AI robot for Mars mission.

- Welcome back to "Yahoo Finance Live. Deeper space exploration missions come with their own set of challenges. Not only are instruments farther away, which make the delay in reacting to certain things difficult, but tricky terrain on the Martian surface has made wheel travel less ideal, as well. And that has researchers turning the Boston Dynamics dog-like robot-- you may remember that one from a lot of viral videos-- SPOT, as it's called, for potential solutions.

For more on that partnership, happy to welcome in our next guest here. Ali Agha, who has a plethora of titles here. But I'm just going to go ahead and call him group leader and roboticist at Nasa JPL. And Ali, really what you're focusing in on here in the project you lead is kind of the autonomy around some of these robots, to maybe try and figure out some of the things on their own to make the missions easier. But talk to me about how SPOT and that side of robotics helps.

ALI AGHA: Exactly, yeah. We here work on autonomy and artificial intelligence for robotic platforms. In some sense, you can think of our work as focused on building brains for robots. And typically, these brains are agnostic to specific robotic platforms. We integrate these. We have wheeled rovers. We have legged platforms, as the one you can see from Boston Dynamics. Even drones and flying robots with applications for terrestrial settings, search and rescue, mining and so on, as well as our main goal, which is space exploration.

And among these different locomotion capabilities and different mobility systems, legged robots are one of the most promising ones because specifically for NASA, targeting exploration of Mars surface, moon surface. We don't have roads there. It's all rugged terrain, off-road setting.

And even on Earth, when you have no road conditions, you typically have animals with legs, right? So legs offer much more capable locomotion ability to go over rocks and different extreme environments. And that's why we are very excited to integrate our autonomous solutions with these legged platforms to enable new kinds of missions.

Story continues

- I assume, too, that the cost has maybe become a little bit more complicated here. When you're thinking about SPOT off the shelf, I think what? It's like $75,000 for one of those things. And then you add on what you guys are working on. So how much more does it kind of come out to you, when you retrofit or add the capabilities that you need to kind of help in these missions?

ALI AGHA: Yeah. First of all, cost is coming down very rapidly. These technologies are just at-- these are the first steps in bringing these technologies to everyday life, to different types of missions. So the cost is rapidly going down and we are hoping that the legged platforms that are going to get cheaper and cheaper.

But you're right. At the moment, a based platform would cost something like around 70k or so. And adding AI, and autonomy, and the sensing payload on top of it kind of doubles the price, roughly speaking. And that's for terrestrial applications. Once there is really a mission to send these to Mars or the moon, there's all plethora of new challenges to be resolved, such as making sure thermally or radiation-wise, you make these robots Mars ready or moom ready, which would be a totally different scale of cost and need there.

- I mean, all the time, these videos go viral for, I guess, stoking fears in what the autonomous robot future might look like. So there are people out there who might be watching who might be afraid of the idea of adding autonomy to that SPOT dog.

But in the tests that you guys have been running so far, what have you learned about how it can help, and how maybe some of those fears are overblown? But also, the timeline to actually get these things up there for the next mission. What's it all look like?

ALI AGHA: Yeah, there's always that perception about what will happen with AI growing and being more and more capable. But I think something typically being missed is it's not growing in isolation. As it grows and gets more capable, humans are getting capable, as well. It's kind of part of us. It's part of the system we're building.

And in that sense, I think we see, similar to many other technologies in the last century and decades, the benefits typically are much higher. And you might remember the event a few years ago, the Thailand boys got stuck in a cave. If there was technologies that autonomously we send robots, they exactly pinpoint, this is the location. This is what capability is needed, or how rescue people can get to the exact point to save these boys, the mission could have been much faster. We save more lives, and so on.

And similarly, in mining disasters, after natural disasters, and oil and gas industry, there's a lot of application domains that these systems can make a very positive impact on everyday human life.

And when it comes to the second part of your question on NASA missions, of course, there's a long road ahead. The steps we are taking here are initial steps to demonstrate that when we go to extreme environments such as caves, such as places on Mars that are really interesting science-wise, this system is able to actually autonomously get to those points without us having prior information about the environment.

But when it comes to the time to really create a mission around these, there are other considerations, such as entry descent landing. Can we land these sorts of platforms nearby those caves or destinations of interest? How do we handle radiation in places like the moon or Mars where there's no thick atmosphere to protect from that? And similarly, how do we handle thermal variation? There's extreme temperatures, and a hot side and a cold side.

And those are the kind of things that, down the road, after the proof of concept is finalized, need to be studied before a mission with a legged robot to Mars becomes a reality.

- Yeah, you say humanity's progressing, as well. I don't know. It might just be the smart people in your lab. You might be overestimating how much humanity outside of the lab has progressed here.

But when you look at Elon Musk and what he is doing at Tesla, also similarly last month introduced their own kind of concept idea of a Tesla bot, an autonomous robot, as well, which is interesting because he's been pretty outspoken about some of those fears of a Terminator like future, as well. So I guess he's changed his mind on that.

But when you look at the progress on autonomy and what you guys are working on to have these robots do things that, to your point, would benefit humanity, how far off is that technology from maybe the consumer space where you could go out and buy one of these on your own?

ALI AGHA: Yeah, I think, first on Elon Musk, I would say what they're doing in SpaceX, it's amazing in the sense that the increase in the frequency of launches from private sector, SpaceX, Blue Origin, all other companies is going to basically expedite by far the amount of technologies and opportunities that's going to be there to colonize other planets. And that's an amazing push there, and it's very, very helpful for developing these sorts of technologies and expediting them.

And when it comes to benefits to humanity, I think it is-- in my opinion, the next era is a robotic and AI era, where basically, the AI comes to physical systems, and embodies and tries to help people. We can see already the impact on education. You can see all sorts of different robots that kids can use to learn coding, to help with their education.

We can see slowly the entrance of robots to health care. We can see a direct impact on-- in the COVID era, basically we saw the direct impact, how robots can sometimes isolate and reduce the risk to the patient, doctor, in hospitals. And search and rescue is definitely another very big application domain where, after natural disasters, there's a clear need to send these robots to save lives or make the operations way more efficient for rescue personnel.

- And we've been seeing a growing number of those natural disasters here. We might need more robots out there than were expected to help on the front. But Ali Agha, group leader and roboticist at NASA JPL, appreciate you coming on here to explain it all for us, man. Have a great weekend. Exciting to see all the progress there.

View original post here:

Sending robots to space is the 'first step' to bringing AI to everday life: Group Leader and Technologist at NASA JPL and Caltech - Yahoo Finance