Robust Robot Placement and Tool Geometry Optimization for Continuous Cutting and Welding Trajectories Under Positioning Uncertainty

This study introduces a computational methodology to determine the optimal robot base placement and tool geometry for welding and cutting applications. Relying on a particle swarm optimization (PSO) algorithm, the approach efficiently explores the configuration space to identify appropriate positions for trajectory execution—providing results significantly faster than traditional grid-based sampling, which linearly evolves with the number of parameters. It is important in the context of welding and cutting task where each use case involves a specific robot position and tool depending on the trajectory and workpiece geometry. To delineate feasibility zones, an α-shape algorithm is used, followed by a Voronoi-based calculation of the largest inscribed circle, used here as a robustness metric to account for placement inaccuracies, such as when robots are deployed on mobile bases. The method integrates an inverse kinematics model to exhaustively evaluate initial configurations and guides the final effector along a reference path using the Jacobian matrix. Each attempt is scored according to the quality of execution, taking into account common boundaries, singularities and workspace limitations. The effectiveness of the methodology is demonstrated with two robotic platforms: the Fanuc CRX10iA/L and WeezLight, a 6-R collaborative robot developed by Weez-U Welding.