Mobile Radiation Source Interception by Aerial Robot Swarms

This paper presents a motion planning approach to track a moving target for the purpose of classifying its radioactivity properties, using a team of micro aerial vehicles (MAV). The team falls into an optimal, from a sensing perspective, formation around the target and collectively decide, within seconds, if the target is even weakly radioactive. The MAV formation is optimal in the sense that the vehicles minimize the detection error in their collective decision-making. Intuition from the analytical solution of the basic problem formulation, coupled with numerical simulations, guides the technical approach in which an intractable nonlinear optimization problem is converted into a quadratic program (QP). The QP solution then informs a motion planner based on navigation functions, to command the MAV into the desired formation. The complete motion planning and control architecture is tested in simulation.