Complete Coverage Path Planning Using Adaptive GBNN for Omnidirectional Sweeping Robot

Outdoor cleaning while maximizing the area coverage is a complex task because of confined spaces, precision motion, and planning challenges. This paper proposes a novel design of an omnidirectional self-reconfigurable robot named PantheraV3. The novelty here is in terms of the ability of the sweeping brushes to rotate and brush along its length and width, resulting mainly in two states of the robot during the cleaning. This paper proposes the Complete Coverage Path Planning (CCPP) method by adapting Glasius Bio-inspired Neural Network (GBNN) to the robot form factor for more efficient area coverage. The algorithm is made generic which can support any form factor or footprint of robots. Real-world experiments were conducted with PantheraV3. Using the proposed aGBNN and PantheraV3 was able to complete area coverage with lesser path length, i.e., 40% reduction in distance traveled than with GBNN in a selected environment.