Dynamic Coverage Path Planning for Mobile Robots: A Spanning Tree Based Approach with Real-Time Obstacle Avoidance

This paper presents a novel dynamic Coverage Path Planning (CPP) algorithm for mobile robots operating in grid-based workspaces, with an emphasis on real-time obstacle avoidance. The environment is partitioned into major nodes and subcells to facilitate efficient area coverage using a spanning tree-based strategy. The robot, equipped with 360-degree LiDAR sensors, continuously scans its surroundings to detect both static and dynamic obstacles within a predefined range. Dynamic obstacles, assumed to move at the same speed as the robot, are managed through predictive redirection mechanisms. Specifically, the robot adapts its trajectory based on the obstacle's direction of motion - rerouting vertically in response to horizontally moving obstacles and horizontally in response to vertically moving ones. This adaptive strategy ensures complete coverage while minimizing the risk of collisions. Simulation results validate the effectiveness of the proposed algorithm in both static and dynamic scenarios, demonstrating superior performance in coverage efficiency and real-time adaptability compared to existing methods.