Path planning for a robotic arm sand-blasting system

Steel bridges are vulnerable to corrosion and their surfaces have to be de-rusted and repainted regularly. Since the process is complicated, expensive and the removed paints are harmful to human workers' health; the use of an automatic robotic system would be an attractive alternative. This paper presents an approach for planning paths for a robotic arm used in the sand-blasting operation. A hexagonal topologybased coverage pattern is adopted to reduce the amount of un-blasted areas and an editing process is included to confine the blasted areas within desirable boundaries. Furthermore, a genetic algorithm is employed to obtain an effective path with minimum arm travel distances and magnitude of turns. Collisions to obstacles are alleviated by making use of the force-field strategy. The effectiveness of the proposed methods is verified by simulations based on an industrial robot arm model and a complex bridge environment. © 2008 IEEE