|Service robots aim at helping humans in their tasks at home or office. Some tasks can be performed by a team of robots that need to visit target locations in the environment. This problem can be modeled as an instance of the multiple Traveling Salesman Problem (mTSP). The goal of mTSP is to minimize the total distance traveled. However, when the total distance is minimized, some salesmen tend to travel more than others depending on the distribution of the cities. Therefore, balancing individual routes is important to ensure equal battery consumption for the team to complete the task. We proposed a centralized system supported by Multi Objective Evolutionary Algorithms (MOEAs) that generate routes by minimizing the sum of travel distance and standard deviation simultaneously. Experiments were carried out to find MOEAs capable of generation high quality routes in a reasonable amount of time. Moreover, we introduce a novel robot-server architecture that connects robots with our system by means of threads, ROS nodes, and Websocket.|
*** Title, author list and abstract as seen in the Camera-Ready version of the paper that was provided to Conference Committee. Small changes that may have occurred during processing by Springer may not appear in this window.