Toward Robotic Triage: a Distributed Task and Motion Planning Framework for Efficient Human-Robot Emergency Response

This paper proposes a novel human-robot coordination framework to enhance medical assistance during urgent disaster relief operations. Specifically, we propose to leverage a heterogeneous robotic system in which multiple UGVs and a UAV seamlessly coordinate with a medic to find and assist victims. The UGVs are tasked with exploring the environment, finding victims, performing basic triage operations, and reporting to a UAV. The UAV operates as a relay to quickly inform the medic about the location and status of victims, while providing an optimized route to follow. A task and motion planning (TAMP) approach is proposed to coordinate UGVs and the UAV with the medic. The problem is then cast as a Traveling Salesman Problem (TSP) incorporating medical policies from domain experts to define the best route for the medic to follow considering victims’ locations and their severity. Extensive simulations and experiments are performed to showcase the effectiveness of the proposed framework compared to different strategies in terms of task allocation and planning. The results of the proposed humanrobot triage approach shows that by strategically coordinating heterogeneous robotic systems it is possible to reduce the amount of time it takes to locate and assist casualties by an expert medic. Note—Simulations and experiment videos can be found in https://www.bezzorobotics.com/sieds25