Optimizing task allocation for multi-robot systems in smart warehouses

Abstract

This article presents a novel framework for Multi-Robot Task Allocation (MRTA) in smart warehouses by using metaheuristic approaches, namely Multi-Objective Variable Neighborhood Search with Adaptive Intensification and Diversification (MOVNS-AINS) and Non-dominated Sorting Genetic Algorithm II (NSGA-II), to optimize task distribution. The proposed approach addresses the challenge of coordinating multiple mobile robots in warehouse logistics, where tasks such as item transportation, storage, and delivery must be continuously assigned under spatial and temporal constraints. Traditional heuristic methods focus on single-objective optimization. Besides, they often struggle to maintain solution quality when the number of tasks and agents increases or when the environment changes dynamically. In contrast, we propose an MRTA optimization framework to address task allocation in structured warehouse environments with heterogeneous robots and dynamic constraints with a multi-objective strategy considering travel distance, workload distribution, and execution time. The proposed framework is validated in a simulated warehouse environment using NVIDIA Omniverse Isaac Sim, demonstrating scalability and adaptability to dynamic task assignments in industrial logistics and manufacturing environments.