Effect of backlash hysteresis of surgical tool bending joints on task performance in teleoperated flexible endoscopic robot

Abstract

Background: The tendon-sheath mechanism provides flexibility but degrades the task performance of the flexible endoscopic robot because of the inherent backlash hysteresis problem. Previous studies have only focused on reducing backlash hysteresis.The goal of this study is to identify the backlash hysteresis criteria of surgical tool bending joints to maintain efficient surgical performance. Methods:A test platform for a surgical tool has been developed that has initial backlash hysteresis under 5 and can adjust the backlash hysteresis intentionally. Performance variation has been investigated in three bench-top endoscopic tasks in which various backlash hysteresis conditions were intentionally adjusted. Results:A clear drop-off in task performance has been observed when the backlash hysteresis of the bending joints was greater than 10 regardless of the type of task and link length. Conclusions:The backlash hysteresis of surgical tool bending joints should be reduced to at least 10 to maintain efficient performance in robotic endoscopic surgery. K E Y W O R D Sendoscopy, force control, human-machine interfaces, telesurgery

Advancement of Flexible Robot Technologies for Endoluminal Surgeries

“…While a TSM-based manipulator has many advantages, its precise control is difficult due to the hysteresis caused by nonlinear friction between the tendon and sheath, backlash, or a dead zone caused by tendon slackening and elongation. This contributes to the delay and degradation of control accuracy in conventional kinematic control models, limiting the performance of flexible surgical robots [46]. Flexible surgical robots for endoluminal applications are prone to hysteresis because their manipulators usually have a long and tortuous tendon configuration, leading to large friction force and wire deformation.…”

Automating Surgical Peg Transfer: Calibration with Deep Learning Can Exceed Speed, Accuracy, and Consistency of Humans

“…Surgical tasks often require positional accuracy of the endeffector in the workspace within 2 mm, and this is difficult to autonomously obtain with cable-driven surgical arms 50,51 due to effects such as cable tension, cable stretch, and hysteresis, 52,53 and such effects may be exacerbated with flexible arms 54 and usage-related wear. To compensate for these errors, prior methods use techniques such as unscented Kalman filters to improve joint angle estimation 8 by estimating cable stretch and friction, 7 or by learning corrective offsets for robot end-effector positions and orientations.…”

Efficiently Calibrating Cable-Driven Surgical Robots with RGBD Fiducial Sensing and Recurrent Neural Networks

“…These surgical tasks often require positional accuracies bounded within 2-3mm, which is difficult to obtain with cabledriven Robot Surgical Assistants such as the da Vinci [10], the Raven II [6], and especially flexible surgical robots [8], [12], as they are known to suffer from cable stretch and tension, backlash, and hysteresis. Research groups have taken various approaches to compensating for these inaccuracies, such as by using Unscented Kalman Filters to improve joint angle estimation [5] estimating cable stretch and friction [15], or directly learning offsets to correct for robot end-effector positions and orientations [17], [13], [25].…”