Cooperative Tracking of Moving Targets by Teams of Autonomous Unmanned Air Vehicles

Teams of low cost unmanned aerial vehicles (UAVs) with autonomous behaviors will be capable of performing inexpensive, persistent distributed sensing functions. One application of interest is the use of UAV teams to perform Cooperative Search, Acquisition and Tracking (CSAT) of moving ground targets in a stealthy manner. Current state-of-the-art for autonomous control of UAV teams enables the collaborative tracking of friendly ground vehicles, for example in the context of a convoy protection scenario. The focus of this work is to demonstrate the joint capability of the MLB Company's small autonomous UAVs and UC Berkeley's self-directed collaborative control architecture to perform CSAT of multiple moving ground or sea-based targets, drawing heavily on field experience by both groups in the areas of UAV operations, ground vehicle tracking, vision-based road following, and collaborative team control. The extension of collaborative tracking of friendly vehicles to CSAT of adversarial targets will be achieved by addressing two major technical challenges. First, stand-off line of sight trajectory following will be integrated with the current collaborative tracking methodology. Second, a data fusion center will be developed that integrates measurements from all team members in order to estimate ground target tracks in a shared global reference frame.