Method for qualitative evaluation of robotic trajectories based on geometric and kinematic metrics

This paper presents a quantitative framework for the qualitative evaluation of robotic trajectories based on geometric and kinematic metrics. While trajectory optimization in industrial robotics is often judged visually or through execution time, such criteria may overlook aspects of motion smoothness and stability. Starting from a set of ten randomly distributed target points, several alternative trajectories with similar total lengths were generated using cubic spline interpolation through different intermediate points. For each trajectory, specific evaluation metrics were computed, including total path length, mean and maximum curvature, jerk-based smoothness indices, and a composite quality score. The obtained values were compared and analyzed to identify correlations between the proposed metrics and the robot’s simulated motion in RoboDK. Results show that even trajectories of comparable length can exhibit significantly different curvature and jerk characteristics, directly influencing execution quality. The proposed metric-based approach enables an objective assessment of trajectory quality and provides a foundation for advanced optimization strategies grounded in physical or data-driven principles.