Maximizing Reputation for Serving Time-Constrained Tasks with Distributed Heterogeneous Teams

This paper presents a distributed approach to dynamic task allocation and scheduling of heterogeneous tasks with multi-robot teams. We consider a heterogeneous team responsible for serving discrete tasks in a convex environment, with each robot equipped with an arbitrary set of capabilities, and each task specified by a due time and processing time. We propose a distributed approach which leverages heterogeneous Voronoi-based coverage control for task allocation, and an Ant Colony System (ACS) for task service sequencing. As each robot completes a task, it develops a reputation for its ability to serve tasks in a timely manner. Robots use this reputation metric to modify their coverage area and consequently the tasks assigned to them. Our approach offers a robust solution to the variations in tasks and team performance over time. We perform simulations against baseline comparison algorithms to show that our approach achieves comparable performance to a state-of-the-art centralized approach.