SAM: Semi-Active Mechanism for Extensible Continuum Manipulator and Real-time Hysteresis Compensation Control Algorithm

TL;DR

This work tackles limited workspace and control accuracy in cable-driven continuum manipulators for minimally invasive surgery by introducing a Semi-active Mechanism (SAM) that extends the workspace without extra mechanical elements. It couples this hardware with a real-time hysteresis compensation strategy based on Temporal Convolutional Networks (TCNs) trained from RGBD fiducial marker data, achieving latency as low as 1 ms. The approach yields a substantial increase in reachable workspace (up to a 527.6% gain) and reduces hysteresis errors in both joint-space and operational-space tasks, with notable improvements on unseen translations and in a box-pointing task. The results indicate strong potential to improve surgical task performance, while acknowledging marker-induced deflection and pointing toward future markerless sensing improvements. The methodology blends physics-based kinematics with data-driven hysteresis modeling to provide robust, real-time control for an extensible CDCM.

Abstract

Cable-Driven Continuum Manipulators (CDCMs) enable scar-free procedures but face limitations in workspace and control accuracy due to hysteresis. We introduce an extensible CDCM with a Semi-active Mechanism (SAM) and develop a real-time hysteresis compensation control algorithm using a Temporal Convolutional Network (TCN) based on data collected from fiducial markers and RGBD sensing. Performance validation shows the proposed controller significantly reduces hysteresis by up to 69.5% in random trajectory tracking test and approximately 26% in the box pointing task. The SAM mechanism enables access to various lesions without damaging surrounding tissues. The proposed controller with TCN-based compensation effectively predicts hysteresis behavior and minimizes position and joint angle errors in real-time, which has the potential to enhance surgical task performance.

Figures (19)

• Figure 1: Concept Design of the Proposed Mechanism: By utilizing the Semi-active segment, the SAM manipulator can access to the target lesions.
• Figure 2: Comparison of Workspace during Translation between Conventional Continuum and Semi-active Continuum: (a) Conventional CDCM with fixed-length segments and a passive flexible segment. This design limits the workspace regardless of manipulator translation and restricts accessibility to the target region. (b) The proposed manipulator with a Semi-active segment that lengthens during translation through the driving part, resulting in an increased workspace within its operational space.
• Figure 3: Extension Principles and Features of the Semi-active Segment : The Semi-active segment integrates both active and passive parts. The active part is directly driven to enable bending motion, while the passive part is flexible and transmits the driving force. As the translation length increases, the active part extends. A key feature of the Semi-active segment is its ability to maintain a consistent bending angle during this extension. The segment maintains a consistent bending angle of 90 degrees during extension: (a) At a translation length of 0 mm. (b) At a translation length of 23 mm. (c) At a translation length of 46 mm.
• Figure 4: Components of the Proposed Continuum Manipulator : Overall structure showing the main components including extensible segment (Semi-active segment), segment 2, forceps, and the axial DOFs of the base.
• Figure 5: Kinematics Diagram of Continuum Segment and the Proposed Manipulator : (a) The coordinate frame and terminology of the continuum segment, (b) The coordinate frame and terminology of the proposed manipulator.