CAREER: Theory and Practice of Autonomous Soaring for Aerial Robots

Small and mini uninhabited aerial vehicles (UAVs) are ideal for many missions which are too risky or infeasible for larger vehicles. However, small size results in performance limitations: there is reduced capacity for on-board energy storage and sensing payload, and viscous forces are a large contributor to drag, making design of efficient vehicles difficult. Birds and insects are faced with similar problems, and they have evolved flight strategies to overcome the challenges associated with small size.

This research aims to develop planning and control algorithms that will enable small and mini UAVs to extract energy from the surrounding atmosphere: like birds and human glider pilots, they will be able to soar. There are three aims: (1) the development of planning algorithms that can account for the complex effects of vehicle kinematics/dynamics and environmental factors such as wind on the feasibility and costs of a trajectory; (2) the development of a vehicle that uses intelligent planning and control algorithms to overcome the challenges associated with its physical constraints; (3) enabling novel science missions. Further, the planning techniques developed here will be applicable to other instances of robot planning under complex dynamic and environmental conditions.

The ultimate goal of this research is to enable a robotic aircraft to track a migrating raptor over extended distances and time. Recognizing that this is an ambitious goal, requiring advances in other areas of robotics research as well as intelligent planning and control, this research identifies several science missions that are enabled in the short and medium term by a soaring-capable mini-UAV. These science missions will be undertaken in collaboration with researchers at Penn State's meteorology department and researchers at the Acopian Center for Conservation Learning at Hawk Mountain Sanctuary.