Coverage Area-Based Weighted Mean Yielding Position Exploration Method for Resolving Deadlocks Among Autonomous Robots

When introducing multiple autonomous service robots, collision avoidance, and deadlock resolution among robots are crucial for efficient navigation. In situations where robots are entirely deadlocked, assigning priority among robots and performing yielding maneuvers is necessary to allow other robots to exit. Previously, for such yielding actions, static waiting positions were pre-set, or collision-free avoidance positions were dynamically calculated based on the relative speeds of robots. However, pre-setting, fixed waiting positions could lead to increased installer fatigue as it depends on the spatial structure of the operating environment, and considering the other robot's speed was ineffective in narrow and long corridors. This paper introduces an algorithm that identifies potential yielding positions within the coverage area and dynamically generates an appropriate yielding position by considering their weighted average. This reduces the need for manual configuration of static positions and ensures effective yielding maneuvers, even in long, narrow corridors where only one robot can navigate.