Autonomous Service Robots for Urban Waste Management - Multiagent Route Planning and Cooperative Operation

Abstract

A sustainable approach towards smart waste management means reduced impact on the health of the workers, lower greenhouse gas emissions and low operational costs. Incorporating fleets of the robot MARBLE (Mobile Autonomous RoBot for Litter Emptying) effectuates these requirements for autonomously emptying the dustbins on the streets of Berlin. The dedicated waypoints for the route formulation are the global positions of the dustbins, and the garbage emptying position from the robots after reaching the maximum compressed-garbage storage capacity. In this paper, we provide a cost function efficient solution for providing global routes to the robots. The cost function includes operational energy expenditure. After conducting numerous experiments with varying algorithm parameter values, the specific weightage of route permutation operators and the initial temperature for a simulated annealing method with nearest neighbour approach was narrowed down as a well-performing set. This set was used in three different operation scenarios. The results of the cost function on the basis of the chosen parameters show an improvement of average 28% for two different exemplary datasets, and 30% improvement for an implementation in an actual existing park in Berlin for operation scenarios as compared to results from a nearest neighbour heuristic.