A Multi-Objective Pareto-Optimal Solution to the Box-Pushing Problem by Mobile Robots

Abstract

The paper provides a new formulation of the well-known box-pushing problem by robots as a multi-objective optimization problem, and presents Pareto-optimal solutions to the problem. The proposed method allows both turning and translation of the box, while shifting it to a desired goal position. Local planning scheme is employed here to determine the magnitude of the forces applied by two mobile robots at specific locations on the box to align and translate it in each distinct step of motion of the box, so as to minimize the consumption of both time and energy. This is realized using non-dominated sorting genetic algorithm-II (NSGA-II). The proposed scheme, to the best of the authors' knowledge, is a first successful communication-free, centralized co-operation between two robots applied in box-shifting, satisfying multiple objectives simultaneously using evolutionary algorithm.