Pith. sign in

REVIEW 3 major objections 4 minor 49 references

A Disturbance in the Force: Force Actuation on the RAVEN II Surgical Robot with Parallel Motor-Cable Units

T0 review · 3 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

Pith's one-line read A six-cable rig applies external forces to a surgical robot's end-effector with sub-newton errors, giving learning-based force estimators the labeled training data they need.

desk verdict Useful integrated hardware system, but the sub-1 N accuracy claim is unproven because the validation uses the same load-cell-based force estimate the controller tracks. read the letter →

arxiv 2608.06488 v1 pith:HMXK7WKN submitted 2026-08-06 cs.RO

classification cs.RO
keywords forceactuationsurgicalroboticsRAVENIIcable-drivenparallelrobotestimationtrainingdatahapticfeedbacksensorlessSLSQPtensionoptimization
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper reports a hardware-software system that pulls on the end-effector of a RAVEN II surgical robot with six motorized cables arranged around the workspace. The system is designed to apply external forces of arbitrary direction and magnitude while the robot moves, so that recorded trajectories carry known force labels. In a preliminary dataset, the applied force differed from the desired force by mean absolute errors of 0.80 N, 0.79 N, and 0.44 N in X, Y, and Z. The motivation is to provide training data for learning-based force estimators that currently lack representative ground-truth force labels. If the accuracy holds, the system offers a practical way to collect such data without adding sensors to the end-effector.

What carries the argument

The load-bearing object is the parallel motor-cable unit: a DC motor driving a cable reel, a load cell sensing cable tension, and a fixing frame that orients the load cell toward the workspace center. Six such units surround the robot, and their tensions combine at the end-effector into a single resultant force. The control loop computes each cable's direction from the motor locations and current end-effector pose, uses SLSQP optimization to find tensions that realize the desired force while maintaining a minimum tension, and closes two feedback loops, one on force error and one on per-cable tension error. A simulation with robot-link bounding boxes and Monte-Carlo workspace sampling is used to choose motor locations that keep tensions within limits and avoid cable interference, and a rigid-alignment algorithm converts measured motor-unit positions into the robot coordinate frame.

What would settle it

Attach a calibrated six-axis force sensor at the RAVEN II end-effector, run the same commanded trajectories with the cable system active, and compare the sensor-measured end-effector force with the commanded force in each direction. Sub-newton mean absolute error on the sensor would confirm the claim; discrepancies above 1 N would show that the tension-and-direction model misses part of the applied force.

Watch

Extended reading notes

Core claim

The central claim is that a parallel motor-cable force actuation system can realize a desired external force on the RAVEN II end-effector with errors below 1 N while the robot follows a trajectory. The authors demonstrate this through a control pipeline: cable directions are computed from motor locations and the current robot state, SLSQP optimization converts the force command into desired cable tensions, and two feedback loops regulate the applied force and individual cable tensions. Validation consists of recorded trajectories with simultaneous position and force commands, and statistical comparison of the commanded force with the realized force gives mean absolute errors under one newton in each coordinate direction. The claim matters because learning-based force estimation needs representative training data in which the robot moves through its workspace under known external forces, which this system intends to supply.

Load-bearing premise

The reported accuracy assumes the force applied at the end-effector is fully determined by the load-cell tension readings and the computed cable directions, without an independent force sensor at the end-effector to confirm that resultant force.

Editorial extensions

If this is right

  • A recorded dataset of RAVEN II trajectories with known external force labels can be built without any end-effector force sensor.
  • Learning-based force estimators trained on such data could estimate contact forces from robot state alone, addressing the missing-haptic-feedback problem on cable-driven surgical robots.
  • The same control pipeline can be reused on other cable-driven robots by updating motor locations and the robot kinematic model.
  • The system can also generate static force profiles at fixed robot poses, useful for studying steady-state force estimation.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the reported sub-newton errors are validated with an independent end-effector force sensor, the same rig could serve as a force-injection ground-truth generator for comparing estimators across different surgical robot platforms.
  • A testable extension is to use the recorded force-labeled trajectories as supervisory targets for a neural network and check whether the resulting sensorless force estimates inherit the sub-newton accuracy.
  • One implicit limit is that the accuracy depends on the cable-direction model; larger robot motion may increase direction error, so the system likely needs periodic re-localization or online calibration.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper presents a parallel motor-cable force actuation system for the RAVEN II surgical robot. Six motor-cable units with load-cell-instrumented cables are installed around the robot workspace, and a two-level controller (higher-level force loop, lower-level tension loop) drives the cable tensions to apply a desired external force to the end-effector. The authors report hardware design, electronics, a simulation for motor placement, a MicroScribe/Kabsch localization procedure, and a preliminary experimental dataset in which the robot follows trajectories under commanded forces. The headline claim is that the actuation system applies force with mean absolute errors of 0.80 N, 0.79 N, and 0.44 N in the X, Y, and Z directions.

Significance. If the sub-1 N accuracy claim were independently established, the system would provide a practical way to generate force-labeled trajectories for learning-based sensorless force estimation, a recognized bottleneck in surgical robotics. The hardware architecture is plausible, the system is described in enough detail to be reproduced, and the authors are appropriately cautious in calling the results preliminary. The main value of the paper, however, depends on the credibility of the reported error metric, and that credibility is currently undermined by a circular validation: the reported 'actual applied force' is computed from the same load-cell measurements and cable-direction model that are used inside the feedback controller.

major comments (3)
  1. [Section II.B / III] The reported accuracy metric is circular. The 'actual applied force' used in computing the mean absolute errors is derived from load-cell tension readings combined with the computed cable directions, which is exactly the same inferred force used by the higher-level feedback controller to compute the error between desired and actual force. Consequently, the experiment measures how well the controller tracks its own internal force estimate, not how accurately the physical force applied to the end-effector matches the command. Any systematic bias in the load-cell calibration, in the straight-line cable-direction assumption, in the alignment of the load-cell fixing frames (Section II.A), or in the motor-unit localization (Section II.E) will be invisible in this error metric. An independent ground-truth check is required—for example, a separate force/torque sensor mounted at the end-effector, or a calibrated spring/mass reference—before the sub-1 N claim can be accepted.
  2. [Section III] The statistical reporting is insufficient to support the headline accuracy claim. Only three mean absolute error values are reported, with no standard deviations, no number of trials or time samples, no maximum errors, and no confidence intervals. The dataset includes both moving trajectory segments and static 'star' segments, but the errors are not broken down by condition. Without these quantities, the reader cannot assess the consistency of the errors, worst-case behavior, or whether a few favorable trajectories dominate the averages. Please report the full distribution of errors, including outliers, and separate static and dynamic cases.
  3. [Section II.A / II.B] The cable-direction model assumes the cable from each motor unit to the end-effector is a straight line defined by the measured motor-unit location and the current end-effector position. This ignores cable sag, elasticity, and any friction or bending at the cable exit or attachment points. The load-cell fixing frames are described as being aligned to the center of the robot workspace, but the end-effector moves throughout the workspace, so the load-cell measurement axis may not remain aligned with the actual cable direction. The paper does not quantify the magnitude of these effects or provide a calibration that accounts for them. Given that the claimed errors are below 1 N, a sensitivity analysis or an experimental validation of the cable-direction model is needed.
minor comments (4)
  1. [Section I] The sentence 'Preliminary experiments suggests' uses a singular verb with a plural subject; change to 'suggest'.
  2. [Section I] The robot name is typeset inconsistently as 'RA VEN-II' and 'RAVEN-II'; please use a single consistent spelling and fix the spacing.
  3. [Section II.B] The paragraph describing the control loop uses numbered steps without a closing punctuation after step 3; add a period and ensure all list items are punctuated consistently.
  4. [Section III] The reader must assume what 'actual applied force' means mathematically; please state explicitly that it is the force computed from load-cell tensions and cable directions, and define the error norm used (e.g., per-axis mean absolute error over all samples).

Circularity Check

1 steps flagged · score 6.0 of 10

Reported sub-1 N force accuracy is the controller's residual against its own load-cell-derived force estimate, not an independent measurement.

  1. self definitional [Section II.B; Section III (Preliminary Results)]
    "There are 2 levels of feedback control. The higher-level feedback control is on the force command, based on the difference between the desired force and the actual applied force. ... Statistical analysis suggested that compared to the desired external force, the actual applied force had mean absolute errors of 0.80 N, 0.79 N, and 0.44 N in X, Y , and Z direction, respectively."

    The 'actual applied force' used in the Section III error computation is not an independent measurement of the force at the end-effector. Per Section II.B, the only force feedback in the system is 'the actual applied force' entering the higher-level feedback loop, computed from the load-cell tensions and the computed cable directions; the lower loop controls the same load-cell tensions. Thus the reported MAE (0.80/0.79/0.44 N) is the residual of the controller tracking its own internal force estimate. The mapping from tension to end-effector force—load-cell calibration, straight-line cable model, no sag/elasticity/friction, load-cell frame alignment, motor localization—is never validated by an independent force sensor.

full rationale

The central accuracy result (Sec. III) is the only quantitative support for the abstract's 'errors less than 1 N' claim. The quantity compared against the command is called 'the actual applied force,' but the paper describes no end-effector force sensor and no independent verification of the tension-to-force model. The control architecture (Sec. II.B) closes a high-level loop on exactly this modeled force, so the reported MAE is the tracking error of that loop against its own estimate. If the model is biased, the controller drives the modeled force to the command while the physical force differs; the reported error cannot reveal this. Hence the validation reduces by construction to a self-consistency check. This is not a self-citation issue and no external benchmark is used; it is a ground-truth/validation circularity. The engineering design and localization work are independent, and the paper is appropriately labeled preliminary, but the headline accuracy claim lacks independent support. Score 6 reflects partial circularity in the central validation.

Assumptions & free parameters 1 free parameters · 2 assumptions · 0 invented entities

The paper introduces no new physical entities. The main assumptions are about the correctness of the force measurement chain: load cell readings, cable directions from kinematics and calibration, and the resulting force estimate. The only explicit hand-tuned parameter mentioned is the minimum cable tension.

free parameters (1)
  • minimum cable tension = not specified
    A minimum cable tension is maintained at all times for smoothness and stability (Section II.B). The value is a hand-chosen parameter that affects the feasible force range and the accuracy of the system, but it is not reported or varied in the experiments.
assumptions (2)
  • domain assumption The resultant end-effector force can be accurately computed from cable tension measurements and cable directions derived from robot kinematics and motor locations.
    The control loop and the validation both rely on this force mapping (Section II.B). No independent force sensor is used to verify that the computed resultant matches the true external force on the end-effector.
  • domain assumption The Kabsch-based transformation between the MicroScribe measurement frame and the robot frame is correct and remains fixed during experiments.
    The motor locations used for cable direction computation depend on this transformation (Section II.E). Any error in this calibration propagates directly into the force computation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Disturbance in the Force: Force Actuation on the RAVEN II Surgical Robot with Parallel Motor-Cable Units." pith.science (2026). https://pith.science/paper/HMXK7WKN

@misc{pith2026260806488,
  author       = {Pith},
  title        = {Pith review of: A Disturbance in the Force: Force Actuation on the RAVEN II Surgical Robot with Parallel Motor-Cable Units},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HMXK7WKN}},
  note         = {Machine review of arXiv:2608.06488}
}
read the original abstract

Difficulty in haptic feedback for surgical robots has been a long-term problem for decades. In recent years, learning-based force estimation from robot states suggests desirable accuracy without the necessity of extra sensors. However, challenges remain in obtaining representative training data in which the robot moves in the workspace under various external forces. In this work, a parallel motor-cable system is developed. With six motor-cable units installed around the robot workspace, cables with controllable tension connected to the robot end-effector can provide the desired external force without interfering with the movement of the surgical robot. The development of the system includes motor-unit hardware, control software, sensor drivers, simulations, and more. Preliminary experiments suggest an accuracy of force actuation with errors less than 1 N.

Figures

Figures reproduced from arXiv: 2608.06488 by the authors.

Figure 1
Figure 1. The hardware and setup of the force actuation system. There are 6 motor-cable units (marked by red circles) installed [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The control workflow of the force actuation system. [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 5
Figure 5. Recorded robot trajectory with external force. The [PITH_FULL_IMAGE:figures/full_fig_p003_5.png] view at source ↗
Figures from the paper (2 more)
Figure 3
Figure 3. Figure 3: 320 Hz load cell drivers for feedback control on cable [PITH_FULL_IMAGE:figures/full_fig_p003_3.png]
Figure 4
Figure 4. Figure 4: Simulation of the force actuation system with [PITH_FULL_IMAGE:figures/full_fig_p003_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 46 canonical work pages

  1. [1]

    Funke, Isabel and Bodenstedt, Sebastian and Oehme, Florian and von Bechtolsheim, Felix and Weitz, J. Using. Proc. Int. Conf. Med. Image Comput. Comput. Assist. Interv. , pages=. 2019 , organization=

  2. [2]

    arXiv preprint arXiv:1412.6980 , year=

    Adam: A method for stochastic optimization , author=. arXiv preprint arXiv:1412.6980 , year=

  3. [3]

    arXiv preprint arXiv:1904.09237 , year=

    On the convergence of adam and beyond , author=. arXiv preprint arXiv:1904.09237 , year=

  4. [4]

    2020 Fourth IEEE International Conference on Robotic Computing (IRC) , pages=

    Collaborative Robotics Toolkit (CRTK): Open software framework for surgical robotics research , author=. 2020 Fourth IEEE International Conference on Robotic Computing (IRC) , pages=. 2020 , organization=

  5. [5]

    2015 IEEE/RSJ international conference on intelligent robots and systems (IROS) , pages=

    Improving position precision of a servo-controlled elastic cable driven surgical robot using unscented kalman filter , author=. 2015 IEEE/RSJ international conference on intelligent robots and systems (IROS) , pages=. 2015 , organization=

  6. [6]

    2016 IEEE International Conference on Robotics and Automation (ICRA) , pages=

    Hysteresis model of longitudinally loaded cable for cable driven robots and identification of the parameters , author=. 2016 IEEE International Conference on Robotics and Automation (ICRA) , pages=. 2016 , organization=

  7. [7]

    IEEE Robotics and Automation Letters , volume=

    Efficiently calibrating cable-driven surgical robots with RGBD fiducial sensing and recurrent neural networks , author=. IEEE Robotics and Automation Letters , volume=. 2020 , publisher=

  8. [8]

    2014 IEEE international conference on automation science and engineering (CASE) , pages=

    Learning accurate kinematic control of cable-driven surgical robots using data cleaning and gaussian process regression , author=. 2014 IEEE international conference on automation science and engineering (CASE) , pages=. 2014 , organization=

Show all 49 references
  1. [9]

    International Journal of Hyperthermia , volume=

    Ablative brain surgery: an overview , author=. International Journal of Hyperthermia , volume=. 2019 , publisher=

  2. [10]

    IEEE Transactions on Systems, Man, and Cybernetics , volume=

    Toward AI research methodology: Three case studies in evaluation , author=. IEEE Transactions on Systems, Man, and Cybernetics , volume=. 1989 , publisher=

  3. [11]

    Speech recognition , pages=

    A tutorial on speech understanding systems , author=. Speech recognition , pages=

  4. [12]

    arXiv preprint arXiv:1901.08644 , year=

    Ablation studies in artificial neural networks , author=. arXiv preprint arXiv:1901.08644 , year=

  5. [13]

    2014 47th Hawaii international conference on system sciences , pages=

    Text simplification tools: Using machine learning to discover features that identify difficult text , author=. 2014 47th Hawaii international conference on system sciences , pages=. 2014 , organization=

  6. [14]

    IEEE Transactions on Automation Science and Engineering , year=

    Automating Surgical Peg Transfer: Calibration With Deep Learning Can Exceed Speed, Accuracy, and Consistency of Humans , author=. IEEE Transactions on Automation Science and Engineering , year=

  7. [15]

    2020 IEEE International Conference on Robotics and Automation (ICRA) , pages=

    Real-time data driven precision estimator for raven-ii surgical robot end effector position , author=. 2020 IEEE International Conference on Robotics and Automation (ICRA) , pages=. 2020 , organization=

  8. [16]

    2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , pages=

    Improving control precision and motion adaptiveness for surgical robot with recurrent neural network , author=. 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , pages=. 2017 , organization=

  9. [17]

    Journal of minimal access surgery , volume=

    Robotic assisted minimally invasive surgery , author=. Journal of minimal access surgery , volume=. 2009 , publisher=

  10. [18]

    Annals of translational medicine , volume=

    Review of robotic-assisted surgery: what the future looks like through a spine oncology lens , author=. Annals of translational medicine , volume=. 2019 , publisher=

  11. [19]

    Surgical endoscopy , volume=

    Review of emerging surgical robotic technology , author=. Surgical endoscopy , volume=. 2018 , publisher=

  12. [20]

    The American Journal of Surgery , volume=

    Robotic technology in surgery: past, present, and future , author=. The American Journal of Surgery , volume=. 2004 , publisher=

  13. [21]

    IEEE Transactions on Biomedical Engineering , volume=

    Raven-II: an open platform for surgical robotics research , author=. IEEE Transactions on Biomedical Engineering , volume=. 2012 , publisher=

  14. [22]

    An open-source research kit for the da Vinci

    Kazanzides, Peter and Chen, Zihan and Deguet, Anton and Fischer, Gregory S and Taylor, Russell H and DiMaio, Simon P , booktitle=. An open-source research kit for the da Vinci. 2014 , organization=

  15. [23]

    2018 IEEE International Conference on Robotics and Automation (ICRA) , pages=

    Fast and reliable autonomous surgical debridement with cable-driven robots using a two-phase calibration procedure , author=. 2018 IEEE International Conference on Robotics and Automation (ICRA) , pages=. 2018 , organization=

  16. [24]

    Mechatronics , volume=

    On the effects of the design of cable-Driven robots on kinematics and dynamics models accuracy , author=. Mechatronics , volume=. 2017 , publisher=

  17. [25]

    2015 IEEE/RSJ international conference on intelligent robots and systems (IROS) , pages=

    Measurement of the cable-pulley Coulomb and viscous friction for a cable-driven surgical robotic system , author=. 2015 IEEE/RSJ international conference on intelligent robots and systems (IROS) , pages=. 2015 , organization=

  18. [26]

    Acta Polytechnica Hungarica , volume=

    The Raven open surgical robotic platforms: A review and prospect , author=. Acta Polytechnica Hungarica , volume=

  19. [27]

    2021 6th IEEE International Conference on Advanced Robotics and Mechatronics (ICARM) , pages=

    A review of bilateral teleoperation control strategies with soft environment , author=. 2021 6th IEEE International Conference on Advanced Robotics and Mechatronics (ICARM) , pages=. 2021 , organization=

  20. [28]

    2015 IEEE International Conference on Robotics and Automation (ICRA) , pages=

    Effects of master-slave tool misalignment in a teleoperated surgical robot , author=. 2015 IEEE International Conference on Robotics and Automation (ICRA) , pages=. 2015 , organization=

  21. [29]

    Otolaryngologic Clinics of North America , volume=

    Past, present, and future of robotic surgery , author=. Otolaryngologic Clinics of North America , volume=. 2020 , publisher=

  22. [30]

    University of Washington, Tech

    Kinematic analysis of the Raven-II research surgical robot platform , author=. University of Washington, Tech. Rep. UWEETR-2012-0006 , year=

  23. [31]

    2017 IEEE International conference on robotics and automation (ICRA) , pages=

    Autonomous suturing via surgical robot: An algorithm for optimal selection of needle diameter, shape, and path , author=. 2017 IEEE International conference on robotics and automation (ICRA) , pages=. 2017 , organization=

  24. [32]

    2014 IEEE International Conference on Robotics and Automation (ICRA) , pages=

    Autonomous multilateral debridement with the raven surgical robot , author=. 2014 IEEE International Conference on Robotics and Automation (ICRA) , pages=. 2014 , organization=

  25. [33]

    2023 International Symposium on Medical Robotics (ISMR) , pages=

    Markerless Suture Needle Tracking From A Robotic Endoscope Based On Deep Learning , author=. 2023 International Symposium on Medical Robotics (ISMR) , pages=. 2023 , organization=

  26. [34]

    2020 29th IEEE international conference on robot and human interactive communication (RO-MAN) , pages=

    Collaborative suturing: A reinforcement learning approach to automate hand-off task in suturing for surgical robots , author=. 2020 29th IEEE international conference on robot and human interactive communication (RO-MAN) , pages=. 2020 , organization=

  27. [35]

    IEEE Transactions on Medical Robotics and Bionics , volume=

    Autonomy for surgical robots: Concepts and paradigms , author=. IEEE Transactions on Medical Robotics and Bionics , volume=. 2019 , publisher=

  28. [36]

    Proceedings of the IEEE , volume=

    Robot-assisted minimally invasive surgery—Surgical robotics in the data age , author=. Proceedings of the IEEE , volume=. 2022 , publisher=

  29. [37]

    Medical Devices: Surgical and Image Guided Technologies , volume=

    Surgical robotics , author=. Medical Devices: Surgical and Image Guided Technologies , volume=

  30. [38]

    2022 , howpublished =

    Hannaford, Blake , title =. 2022 , howpublished =

  31. [39]

    Advances in Mechanical Engineering , volume =

    Sung-Hyun Choi and Jong-Oh Park and Kyoung-Su Park , title =. Advances in Mechanical Engineering , volume =. 2017 , doi =

  32. [40]

    The Encyclopedia of MEDICAL ROBOTICS: Volume 1 Minimally Invasive Surgical Robotics , pages=

    Robot autonomy for surgery , author=. The Encyclopedia of MEDICAL ROBOTICS: Volume 1 Minimally Invasive Surgical Robotics , pages=. 2019 , publisher=

  33. [41]

    2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , pages=

    Surrol: An open-source reinforcement learning centered and dvrk compatible platform for surgical robot learning , author=. 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , pages=. 2021 , organization=

  34. [42]

    Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization , volume=

    SARTRES: A semi-autonomous robot teleoperation environment for surgery , author=. Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization , volume=. 2021 , publisher=

  35. [43]

    2021 IEEE International Conference on Robotics and Automation (ICRA) , pages=

    Deserts: Delay-tolerant semi-autonomous robot teleoperation for surgery , author=. 2021 IEEE International Conference on Robotics and Automation (ICRA) , pages=. 2021 , organization=

  36. [44]

    2012 IEEE International Conference on Robotics and Automation , pages=

    Application of unscented kalman filter to a cable driven surgical robot: A simulation study , author=. 2012 IEEE International Conference on Robotics and Automation , pages=. 2012 , organization=

  37. [45]

    2017 , school =

    Cable Driven Robots: Hysteretic Cable Stretch, Cable-Pulley Network Friction, Fatigue Life, and Kinematics of Two-Arm Multi Staged Flexible Manipulator , author=. 2017 , school =

  38. [46]

    arXiv preprint arXiv:2408.01604 , year=

    Efficient Data-driven Joint-level Calibration of Cable-driven Surgical Robots , author=. arXiv preprint arXiv:2408.01604 , year=

  39. [47]

    2024 , school=

    Tissue Characterization With Surgical Smart Grasper and Hybrid CNN--GRU Model , author=. 2024 , school=

  40. [48]

    2020 IEEE International Conference on Robotics and Automation (ICRA) , pages=

    Neural network based inverse dynamics identification and external force estimation on the da Vinci Research Kit , author=. 2020 IEEE International Conference on Robotics and Automation (ICRA) , pages=. 2020 , organization=

  41. [49]

    IEEE robotics and automation letters , volume=

    Gaussian process regression for sensorless grip force estimation of cable-driven elongated surgical instruments , author=. IEEE robotics and automation letters , volume=. 2017 , publisher=

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.