Evolutionary path planning for robot assisted part handling in sheet metal bending

While the bending sequence planning has been intensively studied, design of the motion path of a sheet metal part in the bending operation tends to be ignored by researchers. Because during the bending operation, the space for maneuvering a sheet metal part is very small, collisions between the part and bending tools are likely to occur. When a robot is used to handle the part, the role of an automatic path-planning tool becomes more significant. In this study, an evolutionary path-planning approach for robot-assisted handling of sheet metal parts in bending is firstly proposed and implemented. The proposed approach globally searches the motion path space to identify feasible paths. Collision detection algorithms based on segment intersection are used to check if the generated paths are feasible or infeasible. This method can automatically design feasible handling operations for a robot. Simulation examples on a simple “V” shaped part and a part with multiple bents demonstrate that the approach is efficient and practical.