Human-Autonomy Teaming-Based Robot Informative Path Planning and Mapping Algorithms with Tree Search Mechanism

Human-autonomy teaming (HAT)-based informative path planning (IPP) approaches are proposed to enable an autonomous robot to explore hazardous environments more efficiently in this paper. The proposed informative path planning approaches based on multi-objective optimization enable the robot to explore the environments, while also visiting several high-interest sites more frequently. Furthermore, the robot can be tailored to human needs; for instance, it reaches the human-set destination point or the direction indicated. One proposed HAT-based IPP method utilizes Rapid-exploration Random Tree (RRT*) to select the most informative tree nodes while traveling to human-set target points within global path planning. The other utilizes the direction set by humans, converts it into direction information. Thus, HAT is integrated in the information map for multi-target Pareto Monte Carlo Tree Search (MCTS) in local path planning. Simulation and comparison studies validate the effectiveness and robustness of the proposed approaches. The results demonstrate that our proposed approaches significantly improve the object search time and trajectory length.