The Mars helicopter made history last April. It was also an exciting time for Stefan Weiss and his staff at the Institute for Intelligent Systems at the University of Klagenfurt: “You are traveling to another planet, you are part of something that has never been experienced and you are excited.”
Journeys to Mars are possible
Even the smallest mistake can lead to the failure of a billion dollar project. But the Mars rover landed as planned. On board was a helicopter with a wingspan of more than a meter. He was supposed to clarify whether flights are possible on Mars, but there are delays: “The takeoff went very well. At first, on the first flight, you could see that some wind had an effect, and the regulator was supposed to act aggressively. Larger against it. If the helicopter is pushed away by the wind, you can quickly return to the target location. This behavior is not set with sufficient force. This basically leads to more stable flight behavior.”
Mars is 480 million km from Earth. The control signal may take eleven minutes to arrive. So the helicopter flies autonomously without any external control: “We have a third of gravity and less than one percent of the atmosphere. That means it’s still quite a challenge to fly something, so the balance is actually not quite the case, but simulations have confirmed severely.”
Children’s carpet simulation
On the floor, tests were carried out with decorative children’s rugs. A camera on the plane takes 30 black and white photos per second. Compared to the following image, the helicopter calculates where it is and where it is heading. The software required for this was developed at the University of Klagenfurter under the direction of Stefan Weiss: “This is very difficult with these optically low-contrast homogeneous regions on Mars. It is difficult to say, is this a stone that we saw earlier or is it a different stone. In this regard, we have strong challenges. Too much compared to children’s rugs. And that’s exactly what led to complications in the sixth flight.”
On the sixth flight, something almost went wrong, creativity faltered, so Weiss: “One assumes it was a certain blur. You thought you could recognize a spot and it wasn’t actually in the right place. And then these misinterpretations lead to wrong behavioral patterns. These are exactly the elements that can lead to a breakdown. Fortunately, this approach is fundamentally strong against such individual errors. But if they occur together more and more often, then the situation, the situation, can certainly be completely wrong, and then what we do not hope It will all happen.”
Single Camera Mounted System
The mission is due to end in August. The professor and his staff are already planning the next steps for more missions and scenarios. The single-camera system remains true: “We want to perceive the entire 3D structure of the environment. Right now, the focus is more on a single level, which means we can’t fly down the crater and look for steep walls yet. You’ll have to upgrade. to a new algorithm. With two cameras, you won’t gain much from flying higher. When the cameras are so close together, it’s like having only one camera on top.”
