Abstract: Autonomous robotic boats equipped with chargers or swappable batteries can serve as Mobile Energy Disseminators (MED) and proactively maintain the State-of-Charge of electric vessels. However, previous methods are either incapable of keeping the vessels continuously driving without recharge downtime, or not directly applicable for the scheduling of chargers on city-scale transportation networks. We propose RoboCharger: a Robotic Charger boat scheduling system that adaptively determines the number of serving MEDs, and the optimal routes of the MEDs according to vessel traffic. We studied a metropolitan-scale vessel mobility dataset provided by MarineTraffic and analyzed the spatial-temporal characteristics of vessel movements in Amsterdam's canals. Based on the analysis insights, we developed a MinHash and spatial-temporal similarity comparison based method for vessel traffic estimation, a Chinese Postman Problem based method for determining the cruising routes of the MEDs, and formulated and solved a multi-objective optimization problem to determine the number of serving MEDs, the driving route of each MED and maintain the SoC of the vessels above zero. Our trace-driven experiments demonstrate that compared with previous methods, RoboCharger increases the average SoC of vessels over all time slots throughout a day by almost 42%, and the number of charges of vessels by almost 53%.
Loading