 For aircraft, weight is everything, especially if you're flying a hybrid unmanned aerial vehicle that weighs just over half a pound. This is Dr. Steve Nogar's daily challenge at the U.S. Army Research Laboratory at Aberdeen Proving Ground, Maryland, where he's trying to create a unique UAV design to give soldiers the advantage on the battlefield of the future. There's a lot of vehicles where to get this type of vehicle, they basically take a quad rotor and basically staple it to a fixed wing aircraft. So it's got four rotors, but then it's got some extra propellers, and it's got all these different actuators. And that's great. It's really easy to design that and control that, but it's not very efficient. You have a lot of wasted weight, and in an aircraft, weight is everything. The Army is researching many drone designs, but this one is special because it can transform in flight. It uses a tilt rotor design. This is kind of like the V-22 Osprey, where the motors tilt themselves. I mean, this is really going to allow these vehicles to better integrate with the soldiers. Soldiers are going to have to be able to interact with these vehicles all the time, and they're going to have to work as a team to achieve their objectives. So you're not going to know ahead of time what you need that vehicle to do. It might need to go over the next hill, see what's there. It might just need to follow above you to kind of give you some scouting. The target size for the UAV is that yellow plastic half circle. Nogar has temporarily attached a bigger paper half circle to the prototype to slow it down during testing. This is our motion capture system, so what it's doing is it's tracking our vehicle in real time, and it's actually using the coordinates to help control the vehicle's position. Building robots includes mastering design concepts, repair of hardware during testing, and a lot of software programming. Nogar said he has spent much of his time coding. I was kind of surprised, but in retrospect, it makes sense. I mean, a robot's a computer, and you have to program a computer. Programming is a basic tool of the roboticist. The next step is continuing to experiment, refine, and experiment more. Takes a lot more work to do the control and study the dynamics of this vehicle. Looking forward, we want to start looking towards how do we perch or land on something in the environment. So what that means is we have to be able to sense the environment. We have to be able to detect a wall or know what's around the vehicle. We have to be able to do that at a very high rate of speed. If you're going to land on something, you need to know very quickly how fast that's coming up to you as you come into land. And so that means you do have to do a lot of this AI. You have to be able to do very fast perception. We can't put a lot of sensors on this vehicle. It's basically what can you get away with with just one camera? This is doing something that's harder, but will definitely benefit the vehicle as a whole once it's finished. For ARL TV, I'm David McNally.