An Effectiveness Study Across Baseline and Learning-based Force Estimation Methods on the da Vinci Research Kit Si System

2024-05-13

AI Summary

The dVRK-Si surgical robot lacks native force sensing, hindering haptic feedback essential for laparoscopic-like teleoperation. Method: We propose a learning-based end-effector force estimation framework to restore tactile perception. First, we systematically characterize dVRK-Si’s dynamic deficiencies—identifying mechanical imbalance-induced failure of gravity compensation as the primary cause of PID control degradation and force estimation bias. Our model integrates learned dynamics corrections with physics-informed constraints. Results: Evaluated on dVRK-Si, our approach achieves a 5.21% average RMSE—matching the accuracy of the predecessor dVRK-Classic but exhibiting a 2–3× performance gap, revealing inherent modeling challenges in the newer platform. This work establishes the first reproducible benchmark for sensorless force estimation on dVRK-Si and introduces a closed-loop evaluation paradigm integrating dynamic diagnostics, model refinement, and control validation.

Technology Category

• Intelligent Robots: State Estimation
• Planning, Routing, and Scheduling: Replanning and Plan Repair
• Machine Learning: Calibration & Uncertainty Quantification

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Robot-assisted minimally invasive surgery, such as through the da Vinci systems, improves precision and patient outcomes. However, da Vinci systems prior to da Vinci 5, lacked direct force-sensing capabilities, forcing surgeons to operate without the haptic feedback they get through laparoscopy. Our prior work restored force sensing through machine learning-based force estimation for the da Vinci Research Kit (dVRK) Classic. This study extends our previous method to the newer dVRK system, the dVRK-Si. Additionally, we benchmark the performance of the learning-based algorithm against baseline methods (which make simplifying assumptions on the torque) to study how the two systems differ. Results show the learning-based method achieves an average root-mean-square-error (RMSE) of 5.21%, for the dVRK-Si, which is comparable to the dVRK Classic. In both systems, the learning-based method outperforms baselines, but the difference is much larger in the dVRK-Si. Nonetheless, dVRK-Si force estimation accuracy lags behind the dVRK Classic, with RMSE 2 to 3 times higher. Further analysis reveals poor PID control in the dVRK-Si. We hypothesize that this is due to the lack of gravity compensation, as unlike the dVRK Classic, the dVRK-Si is not mechanically balanced. This study advances the understanding of learning-based force estimation and is the first work to characterize the dynamics of the new dVRK-Si system.

Problem

• Extends machine learning-based force estimation to dVRK-Si system.
• Benchmarks learning-based methods against baseline force estimation techniques.
• Analyzes dVRK-Si dynamics and identifies gravity compensation issues.

Innovation

• Machine learning-based force estimation
• Benchmarking against baseline methods
• Characterizing dVRK-Si system dynamics