UAS flight operations in research: Beating Murphy, satisfying scientists

Marc Schwarzbach, German Aerospace Center, DLR, Institute of Robotics and Mechatronics, Wessling, Germany
Maximilian Laiacker, German Aerospace Center, DLR, Institute of Robotics and Mechatronics, Wessling, Germany
Konstantin Kondak, German Aerospace Center, DLR, Institute of Robotics and Mechatronics, Wessling, Germany

Abstract

Over the last years, the use of UAS as a research tool at universities and institu-tions has increased dramatically. This is due to the good availability of systems and more flexibility in experiments compared to manned systems. The downside is the lower reliability of the systems since many components are adapted from hobby products.

We present methods of improving research UAS reliability by using higher grade components for certain functions or implementing failure tolerant concepts, depending on aircraft size. The methods have been successfully applied in many flights over the last years.

The “Flying Robots” group of German Aerospace Center (DLR) is implementing robotic technology, in software and hardware, to flying systems. As part of the “Robotics and Mecha-tronics Center (RMC)” we can benefit from more than 20 years of robotic research. For flight experiments, fixed wing and helicopter platforms in a weight range of 10kg to 130kg are used.

The main topics of our research are high performance control, multi-sensor based state esti-mation, aerial manipulation and high altitude platforms. Flight experiments in these fields require tight coupling of new experimental functionalities with the flight control systems. To give the maximum flexibility to the scientists, there should be no complex certification pro-cesses for new functionalities. Unified system components and interfaces for all flying plat-forms and proven preflight acceptance testing procedures provide short time from idea to flight. This is possible by a system design that implements a reliable way to disconnect the experimental flight controller and have the aircraft controlled by a safety pilot on the ground. Additionally, main control components like servo motors and batteries are installed twice.

System monitoring before takeoff and during flight is also very important to detect failures in the experimental and flight components. In contrast to operational autopilot systems we do not limit the displayed values but show all with the most important highlighted and color cod-ed. This allows fast interpretation of unexpected system behavior.

Driven by the demands of a high altitude long endurance mission for the ELHASPA aircraft, our systems were extended to implement a failure tolerant concept. Any failure of one system component will not lead to loss of control and the aircraft can be landed safely. The flexibility described above could however be sustained.