A heterogeneous team of unmanned vehicles is given an operations area with an unknown number of moving target vehicles. The team must search the operations area to find these vehicles and then keep track of their locations. The searching and tracking is coordinated between the team members using a decentralized task assignment algorithm on each vehicle, and without a central planner. Furthermore, because the vehicles are operating at low altitude and communication is limited to line-of-sight, communication between the vehicles is intermittent. A simulation for testing various task assignment algorithms has been implemented.
The CSAT project’s goal is to demonstrate a complete system that involves flight ready hardware, software and operator interface. On the hardware side, the focus is on integration of autonomous aerial vehicles with onboard vision sensor, and onboard computer capable of processing the vision data as well as the execution of the planning algorithm. On the software side, the project will develop algorithms to perform the task of search, detection, tracking, and planning. Further research will focus on implementation of image sensor, the vision software, and proof of concept testing within RAVEN.