Overcoming Imperfect Kinematics in Surgical Robotics Through Sim-to-Real Visuomotor Learning

2026-06-19

AI Summary

This work addresses the challenge of degraded kinematic model accuracy in surgical robots due to unreliable internal sensors, which limits the precision of autonomous control. To overcome this limitation, the authors propose an end-to-end visuomotor control approach based on a teacher–student framework that integrates external visual feedback with inherently noisy internal sensor data. By leveraging simulation-to-reality (sim-to-real) transfer learning, imitation learning, and multimodal sensor fusion, the method achieves closed-loop error compensation without requiring an exact physical model or elaborate calibration procedures. Evaluated on the da Vinci Research Kit, the proposed approach demonstrates substantial improvements in both control accuracy and robustness, offering a novel paradigm for mitigating performance degradation caused by sensor imperfections in robotic surgery systems.

Abstract

Robot-Assisted Surgery is integral to modern minimally invasive procedures, with automation emerging as the next frontier to enhance precision and reduce surgeon fatigue. This evolution is largely impeded by the inherent kinematic inaccuracies of surgical robots, where unreliable internal sensors lead to significant control errors. While previous methods attempted to mitigate these issues through complex model-based calibration, they often suffer from high cost and limited effectiveness. This work utilises a learning-policy to actively compensate for hardware inaccuracies using closed-loop visual feedback that was trained from a teacher-student learning framework. The policy can fuse unreliable internal readings with precise external visual data, allowing it to correct for kinematic errors in real time without needing a perfect physical model. The learned policy was successfully deployed on the da Vinci Research Kit, where experiments validated the fundamental feasibility of using external vision to overcome internal sensor deficits. This research provides a foundational and reliable control methodology, paving the way for more advanced and robust surgical automation.

Problem

• surgical robotics
• kinematic inaccuracies
• sensor reliability
• robot-assisted surgery
• control errors

Innovation

• sim-to-real
• visuomotor learning
• kinematic error compensation
• surgical robotics
• teacher-student learning