Airborne sensor and perception management: Experiments and Results for surveillance UAS

Unmanned Aerial Systems (UAS) will be used in more and more complex surveillance and reconnaissance missions. Such UASs are expected to conduct a variety of different sensor oriented tasks while operating in varying environmental conditions. These assumptions put high requirements on the perception and sensor capabilities of the UAS. In order to cope with these requirements in a more flexible and automated way, we propose a high-level concept for a Sensor and Perception Management System (SPMS). The system receives and computes high level tasks (e.g. observing a determined area) and dynamically reconfigures itself using appropriate algorithmic solutions taking into account available onboard sensory and computational resources as well as domain knowledge and actual environmental information during flight. In this paper we describe implementation aspects and investigate performance and capabilities of our concept in detail, derived from experiments within a civil Search and Rescue (SAR) scenario.