ARGoS Simulator-Based Decentralized Swarm Control for Autonomous Smart Mobility and Logistics Systems in Unknown Environments: A Self-Organizing Potential Field Approach

Smart mobility and logistics ecosystems increasingly rely on fleets of autonomous ground agents (e.g., last-mile delivery robots, automated guided vehicles in warehouses/ports, and cooperative inspection platforms ) that must navigate partially known environments under limited coordination. This paper presents a fully decentralized potential-field control framework implemented in the ARGoS simulator for multi-robot navigation in such settings, using only onboard sensing and local inter-agent communication (no maps or centralized supervision). Each robot independently computes attractive, repulsive, and formation-regulating forces to achieve trajectory tracking, obstacle avoidance, and adaptive formation maintenance. The framework demonstrates scalability (tested across 13–35 robots), robustness to dynamic reconfiguration, and reliable navigation through constrained environments, including narrow passages and wall boundaries. These properties make the proposed method suitable for distributed autonomous mobility and logistics applications where resilience, scalability, and map-free operation are critical.