Kutlu, Mustafa (Military Technological College), Mahmud, Shekhar (Military Technological College), Freeman, Christopher Thomas (University of Southampton)

Iterative Impedance Learning Control with Explainable AI

Scheduled for presentation during the Regular Session "AI-based Robot Control I" (WeAT2), Wednesday, July 16, 2025, 10:40−11:00, Room 106

Joint 10th IFAC Symposium on Mechatronic Systems and 14th Symposium on Robotics, July 15-18, 2025, Paris, France

This information is tentative and subject to change. Compiled on July 16, 2025

Abstract

This paper introduces an innovative framework that combines Iterative Impedance Learning Control (IILC) with Explainable Artificial Intelligence (XAI) to improve the performance and interpretability of robotic systems. The suggested methodology is implemented on a spring-mass-damper system, a fundamental model for robotic interactions with compliant surroundings. The framework enhances trajectory tracking precision and system adaptability by integrating iterative learning techniques with adaptive impedance control. The use of XAI, utilizing SHAP (Shapley Additive Explanations), delivers immediate insights into the impact of control parameters, enhancing user comprehension and confidence. Experimental findings indicate a 52% reduction in root-mean-square error (RMSE) and a 37% drop in settling time across five learning iterations. The explainability layer emphasizes the proportionate gain (Kp) and learning rate (L) as essential factors, consistent with classical control theory and empirical findings. User assessments of the XAI module produced elevated interpretability ratings, validating its efficacy in connecting sophisticated control algorithms with human comprehension.