Decentralized Information-Rich Planning and Hybrid Sensor Fusion for Uncertainty Reduction in Human-Robot Missions

Abstract

This paper introduces a novel planning and estimation framework for maximizing infor- mation collection in missions involving cooperative teams of multiple autonomous vehicles and human agents, such as those used for multi-target search and tracking. The main contribution of this work is the scalable uni cation of e ective algorithms for distributed high-level task planning, decentralized information-based trajectory planning, and hybrid Bayesian information fusion through a common Gaussian mixture uncertainty representa- tion, which can accommodate multiple mission objectives and constraints as well as het- erogeneous human/robot information sources. The proposed framework is validated with promising results on real hardware through a set of experiments involving a human-robot team performing a multi-target search mission.