The Autonomous Recharging Problem: Formulation and a market-based solution

Abstract

As robots are increasingly employed for distributed missions, an important problem that needs to be solved is the Autonomous Recharging Problem (ARP); that is, effectively planning and coordinating when, where, and how to recharge robots to maximize operational efficiency. In this paper, we define the ARP and clearly describe the relevant technical challenges that need to be addressed in order to solve the ARP. These challenges include individual awareness about energy for each robot, correlating task execution to its impact on battery life, designing and implementing static and mobile recharging stations, optimizing individual worker robots schedules to ensure extended operations, and coordinating a team of worker and recharging robots to effectively share available recharging resources. In this paper we also describe an initial implementation of a market-based solution to the ARP that scales with the number of tasks as well as with the number of robots, and schedules workers' tasks to coordinate with recharging tasks. The developed solution has been evaluated in simulation and on Pioneer P3DX mobile robots executing transportation tasks in an indoor environment. Results demonstrate that our approach consistently outperforms the state of the art in recharging strategies.