Anticipatory Planning and Dynamic Lost Person Models for Human-Robot Search and Rescue

In this work, we consider the problem of planning paths for a team of autonomous unmanned aerial vehicles (UAVs) to assist search and rescue practitioners. To address the problem, we develop a fully integrated framework that includes information from all aspects of the search environment. We take into consideration lost person motion via a behavior-based predictive model, anticipated human searcher trajectories, as well as measurements from fixed field of view sensors on board UAVs. We use a metric of posterior risk as the optimization target as it is an indicator of improved situational awareness and the effectiveness of continuing search efforts. Monte Carlo simulations are presented to demonstrate the effectiveness of the proposed framework.