pith. sign in

arxiv: 2606.04157 · v1 · pith:U3O72YUDnew · submitted 2026-06-02 · 💻 cs.RO

Selecting haptic guidance models in teleoperation: guidelines from a comparative user study

Alexis Boulay (AUCTUS) , Margot Vulliez (AUCTUS) , David Daney (AUCTUS) This is my paper

Pith reviewed 2026-06-28 09:48 UTC · model grok-4.3

classification 💻 cs.RO
keywords haptic guidanceteleoperationuser studyspring-damperpotential fieldguiding tubemodel selectionforce feedback
0
0 comments X

The pith

Haptic guidance models perform differently by environment with no universal winner.

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

The paper compares three haptic guidance models in teleoperation by casting them as variants of a stiffness-damping system and testing them in a vertical farming task across six scenarios. Spring-damper guidance performs best in cluttered spaces, potential-field guidance suits open areas but carries collision risks near obstacles, and the guiding-tube approach supplies a workable middle option. New objective metrics are introduced to assess the human-robot interaction, and measured guiding force is shown to track with reported comfort and trust. The results supply concrete rules for picking a model according to environmental features rather than assuming one approach fits every case.

Core claim

The central claim is that haptic guidance in teleoperation has no single best model; performance instead depends on the surrounding conditions, with spring-damper favored in clutter, potential field favored in free space (with noted risks), and guiding tube providing balance, all derived from direct user comparison and new interaction metrics.

What carries the argument

Unified stiffness-damping formulation with model-specific guiding functions, used to run head-to-head comparison of spring-damper, potential field, and guiding tube in the same experimental setup.

If this is right

  • Spring-damper should be selected when the workspace contains many obstacles.
  • Potential-field guidance should be avoided or augmented with safeguards when the end-effector must pass close to obstacles.
  • Guiding-tube behavior offers a stable default when environmental density is unknown in advance.
  • Real-time monitoring of guiding-force magnitude can serve as an online indicator of operator comfort and trust.
  • The proposed objective metrics supply an alternative to purely subjective questionnaires for evaluating haptic interfaces.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same model-ranking logic could be tested in domains such as underwater or aerial teleoperation to check whether clutter versus openness remains the decisive factor.
  • Control software could monitor force magnitude continuously and switch models on the fly when the metric crosses a comfort threshold.
  • The observed force-comfort link implies that designers should prioritize low-magnitude guidance even when task performance metrics appear acceptable.
  • Future work could add dynamic obstacles or multi-operator teams to see whether the current guidelines still hold under changing conditions.

Load-bearing premise

The vertical farming task and its six environmental scenarios are representative enough to support general guidelines for choosing haptic models in other teleoperation settings.

What would settle it

A replication study performed in a markedly different domain, such as remote assembly or surgical manipulation, that yields a different ranking of the three models or no environment-dependent performance gaps would falsify the claimed selection guidelines.

Figures

Figures reproduced from arXiv: 2606.04157 by Alexis Boulay (AUCTUS), David Daney (AUCTUS), Margot Vulliez (AUCTUS).

Figure 1
Figure 1. Figure 1: Overview of the proposed approach: an interaction evaluator is embedded [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: A guiding force Fg is generated on the haptic interface based on the guidance model and the distance between the robot position Xr and a virtual object (goal Xg, path point Xd, or obstacle Xo). Any guiding force can be unified through a stiffness-damping expression, with the stiffness and damping gains Kg ∈ IR3×3 and Bg ∈ IR3×3 , and the velocity of the haptic device X˙ h ∈ IR3 . The behavior of the guidin… view at source ↗
Figure 3
Figure 3. Figure 3: Force profiles for the different guidance models: (a) force is proportional to distance; (b) force is equal to zero until radius dg, then it is proportional; (c) force increases near attraction point, decreases below dg. The forces profiles are represented in static conditions (no damping effect) with kg = 1. The spring-damper model is used for precisely guiding movements in tasks like surgical teleoperati… view at source ↗
Figure 4
Figure 4. Figure 4: Illustration of path-tracking accuracy for spring-damper/guiding tube and [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Experimental setup and scenarios for the user study. (a) The 6 scenarios [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Bilateral teleoperation control diagram. [PITH_FULL_IMAGE:figures/full_fig_p009_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Trajectory of three participants in scenario [PITH_FULL_IMAGE:figures/full_fig_p010_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Participants’ ranking of guidance models across different scenarios. [PITH_FULL_IMAGE:figures/full_fig_p011_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Significant differences are indicated by asterisks (* [PITH_FULL_IMAGE:figures/full_fig_p012_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Completion time across guidance models in all scenarios except [PITH_FULL_IMAGE:figures/full_fig_p012_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Minimum distance to obstacles across guidance models in all scenarios [PITH_FULL_IMAGE:figures/full_fig_p013_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Mean guiding force magnitude (MC and MT1 ) across all scenarios. Sig￾nificant differences are indicated by asterisks (*) and are based on the mean values (black dots) magnitude as a real-time indicator to adapt the guidance model and optimize user comfort is also promising. 5 Conclusion We propose an overview of the main haptic guidance models that can be used in teleoperation tasks. We define a unified f… view at source ↗
read the original abstract

Haptic guidance in teleoperation enhances operator performance through force feedback. This paper presents guidelines to select the most appropriate model considering the task, the environment and the operator. We define a unified formulation expressing most common models (spring-damper, potential field, and guiding tube) as variations of a stiffness-damping system with model-specific guiding functions. We conducted a user study comparing the three classical models across six scenarios with varying environmental conditions in a vertical farming task. Results show no universally superior model: spring-damper excels in cluttered environments, potential field in free spaces (but it shows risks near obstacles), and guiding tube offers a balanced compromise. We propose novel objective metrics to evaluate the interaction, and show that guiding force magnitude correlates with comfort and trust scores. These findings provide practical model selection guidelines through environmental characteristics and real-time evaluation metrics.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The manuscript unifies common haptic guidance models (spring-damper, potential field, guiding tube) as variants of a stiffness-damping system with model-specific guiding functions. It reports a user study comparing the three models across six scenarios in a vertical farming teleoperation task, finding no universally superior model: spring-damper excels in clutter, potential field in free space (with obstacle risks), and guiding tube as a balanced option. Novel objective metrics are introduced, guiding force magnitude is shown to correlate with comfort/trust scores, and the results are used to derive practical model-selection guidelines based on environmental characteristics.

Significance. If the empirical results and correlations prove robust, the work supplies concrete, environment-dependent selection guidelines that could improve safety and performance in teleoperated systems. The unified formulation and proposed metrics are incremental but useful contributions to the haptic guidance literature.

major comments (2)
  1. [Discussion / Conclusion] The central claim that the study yields general 'practical model selection guidelines' for teleoperation rests on the unexamined assumption that the vertical farming task and its six scenarios are representative of broader domains (e.g., surgery, manufacturing). No cross-task validation, argument for representativeness, or discussion of task-specific biases is supplied; this directly affects the scope of the guidelines.
  2. [Abstract / Results] The abstract (and potentially the results section) reports performance rankings and correlations without participant counts, statistical tests, p-values, or error bars. If the main text similarly omits these details, the reliability of claims such as 'spring-damper excels in cluttered environments' and the force-comfort correlation cannot be assessed.
minor comments (1)
  1. [§2 (Formulation)] Notation for the unified stiffness-damping formulation could be clarified with an explicit table mapping each model to its guiding function parameters.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback. We address each major comment below and agree to revisions that strengthen the manuscript's clarity and rigor without overstating the results.

read point-by-point responses

  1. Referee: [Discussion / Conclusion] The central claim that the study yields general 'practical model selection guidelines' for teleoperation rests on the unexamined assumption that the vertical farming task and its six scenarios are representative of broader domains (e.g., surgery, manufacturing). No cross-task validation, argument for representativeness, or discussion of task-specific biases is supplied; this directly affects the scope of the guidelines.

    Authors: We agree that the guidelines are derived from one task domain and that explicit discussion of scope is needed. The vertical farming scenarios were chosen to capture common teleoperation challenges (clutter vs. free space), but we did not provide arguments for representativeness or address potential biases. In revision we will add a paragraph to the Discussion explicitly limiting the guidelines to similar structured environments, noting the absence of cross-domain validation, and recommending future studies in other domains such as surgery or manufacturing. This will qualify the claims appropriately. revision: yes

  2. Referee: [Abstract / Results] The abstract (and potentially the results section) reports performance rankings and correlations without participant counts, statistical tests, p-values, or error bars. If the main text similarly omits these details, the reliability of claims such as 'spring-damper excels in cluttered environments' and the force-comfort correlation cannot be assessed.

    Authors: The main text reports 12 participants, ANOVA results with p-values, and error bars on all figures. The abstract, however, omits these details. We will revise the abstract to state the participant count, note that reported differences reached statistical significance, and indicate that the force-comfort correlation was assessed with appropriate tests. This ensures all claims are presented with the necessary statistical context. revision: yes

Circularity Check

0 steps flagged

Empirical user study exhibits no circular derivations

full rationale

The paper is a comparative user study that unifies three haptic models via a shared stiffness-damping formulation, runs experiments across six scenarios in one vertical farming task, defines objective metrics from the collected data, and reports observed correlations and performance rankings. No load-bearing step reduces a claimed prediction or guideline to a fitted parameter, self-citation, or definitional equivalence; the guidelines are explicitly presented as empirical observations rather than derived necessities. The representativeness assumption noted by the skeptic is a question of external validity, not an internal circular reduction.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The claims rest on the assumption that the user study results are valid and generalizable, with no free parameters explicitly mentioned.

axioms (1)
  • domain assumption Most common haptic guidance models can be expressed as variations of a stiffness-damping system with model-specific guiding functions.
    This is the basis for the unified formulation mentioned in the abstract.

pith-pipeline@v0.9.1-grok · 5684 in / 1209 out tokens · 33372 ms · 2026-06-28T09:48:07.044431+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

22 extracted references · 17 canonical work pages

  1. [1]

    IEEE Trans

    Abi-Farraj, F., Pacchierotti, C., Arenz, O., Neumann, G., Giordano, P.: A haptic shared-control architecture for guided multi-target robotic grasping. IEEE Trans. Haptics pp. 270–285 (2020).https://doi.org/10.1109/TOH.2019.2913643

    work page doi:10.1109/toh.2019.2913643 2020

  2. [2]

    Alberto, N.: Predictive control of collaborative robots in dynamic contexts. Ph.D. thesis, Univ. Bordeaux (2023)

    2023

  3. [3]

    Benhabib, N., Padois, V., Daney, D.: An approach to quantify manual expertise with collaborative robotics in mind (2021), working paper or preprint

    2021

  4. [4]

    In: Progress in Flight Dynamics, Guidance, Navigation, and Con- trol

    Binet, L., Rakotomamonjy, T.: Using haptic feedbacks for obstacle avoidance in helicopter flight. In: Progress in Flight Dynamics, Guidance, Navigation, and Con- trol. pp. 47–70 (2018).https://doi.org/10.1051/eucass/201810047

    work page doi:10.1051/eucass/201810047 2018

  5. [5]

    IEEE Trans

    Bowyer, S., Davies, B., y Baena, F.: Active constraints/virtual fixtures: A survey. IEEE Trans. Robot. pp. 138–157 (2013).https://doi.org/10.1109/TRO.2013. 2283410

    work page doi:10.1109/tro.2013 2013

  6. [6]

    Davies, B., Jakopec, M., Harris, S., Baena, F., Barrett, A., Evangelidis, A., Gomes, P., Henckel, J., Cobb, J.: Active-constraint robotics for surgery. Proc. IEEE pp. 1696–1704 (2006).https://doi.org/10.1109/JPROC.2006.880680

    work page doi:10.1109/jproc.2006.880680 2006

  7. [7]

    In: pHealth, pp

    Georgsson, M.: Nasa rtlx as a novel assessment for determining cognitive load and user acceptance of expert and user-based evaluation methods exemplified through a mhealth diabetes self-management application evaluation. In: pHealth, pp. 185– 190 (2019). https://doi.org/10.3233/978-1-61499-975-1-185

    work page doi:10.3233/978-1-61499-975-1-185 2019

  8. [8]

    IEEE Transactions on Robotics pp

    Jorda, M., Vulliez, M., Khatib, O.: Local autonomy-based haptic-robot interaction with dual-proxy model. IEEE Transactions on Robotics pp. 2943–2961 (2022). https://doi.org/10.1109/TRO.2022.3160053

    work page doi:10.1109/tro.2022.3160053 2022

  9. [9]

    Yoon, H.-S., Hwang, J.-H., Kim, C., Son, E

    Joseph,L.,Pickard,J.,Padois,V.,Daney,D.:Onlinevelocityconstraintadaptation for safe and efficient human-robot workspace sharing. IEEE Robot. Autom. Lett. pp. 11045–11051 (2020).https://doi.org/10.1109/IROS45743.2020.9340961

    work page doi:10.1109/iros45743.2020.9340961 2020

  10. [10]

    In: IEEE/RSJ IROS

    Li, M., Kapoor, A., Taylor, R.: A constrained optimization approach to virtual fixtures. In: IEEE/RSJ IROS. pp. 1408–1413 (2005).https://doi.org/10.1109/ IROS.2005.1545420

    arXiv 2005

  11. [11]

    Muir, B., Moray, N.: Trust in automation. part ii. experimental studies of trust and human intervention in a process control simulation. Ergonomics pp. 429–460 (1996). https://doi.org/10.1080/00140139608964474

    work page doi:10.1080/00140139608964474 1996

  12. [12]

    In: 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp

    Mühlbauer, M., Steinmetz, F., Stulp, F., Hulin, T., Albu-Schäffer, A.: Multi-phase multi-modal haptic teleoperation. In: IEEE/RSJ IROS. pp. 7734–7741 (2022). https://doi.org/10.1109/IROS47612.2022.9981525

    work page doi:10.1109/iros47612.2022.9981525 2022

  13. [13]

    In: ACM/IEEE HRI

    Nikolaidis, S., Zhu, Y., Hsu, D., Srinivasa, S.: Human-robot mutual adaptation in shared autonomy. In: ACM/IEEE HRI. pp. 294–302 (2017).https://doi.org/10. 1145/2909824.3020252

    arXiv 2017

  14. [14]

    In: Proc

    Olsen, D., Goodrich, M.: Metrics for evaluating human-robot interactions. In: Proc. PERMIS (2003)

    2003

  15. [15]

    Prewett, M., Johnson, R., Saboe, K., Elliott, L., Coovert, M.: Managing workload in human-robot interaction: A review of empirical studies. Comput. Hum. Behav. pp. 840–856 (2010).https://doi.org/10.1016/j.chb.2010.03.010

    work page doi:10.1016/j.chb.2010.03.010 2010

  16. [16]

    Raiola, G., Restrepo, S., Chevalier, P., Rodriguez-Ayerbe, P., Lamy, X., Tliba, S., Stulp, F.: Co-manipulation with a library of virtual guiding fixtures. Auton. Robots pp. 1037–1051 (2018).https://doi.org/10.1007/s10514-017-9680-7

    work page doi:10.1007/s10514-017-9680-7 2018

  17. [17]

    In: IEEE VR

    Rosenberg, L.: Virtual fixtures: Perceptual tools for telerobotic manipulation. In: IEEE VR. pp. 76–82 (1993).https://doi.org/10.1109/VRAIS.1993.380795 Selecting haptic guidance models in teleoperation 17

    work page doi:10.1109/vrais.1993.380795 1993

  18. [18]

    In: IEEE RO-MAN

    Shi, C., Madera, J., Boyea, H., Fey, A.: Haptic guidance using a transformer-based surgeon-side trajectory prediction algorithm for robot-assisted surgical training. In: IEEE RO-MAN. pp. 1942–1949 (2023).https://doi.org/10.1109/RO-MAN57019. 2023.10309357

    work page doi:10.1109/ro-man57019 1942

  19. [19]

    Robotica pp

    Sánchez Restrepo, S., Raiola, G., Guerry, J., D’Elia, E., Lamy, X., Sidobre, D.: Toward an intuitive and iterative 6d virtual guide programming framework for assisted human–robot comanipulation. Robotica pp. 1778–1806 (2020).https:// doi.org/10.1017/S0263574720000016

    work page doi:10.1017/s0263574720000016 2020

  20. [20]

    IEEE Trans

    Zanchettin, A., Ceriani, N., Rocco, P., Ding, H., Matthias, B.: Safety in human- robot collaborative manufacturing environments: Metrics and control. IEEE Trans. Autom. Sci. Eng. pp. 882–893 (2016). https://doi.org/10.1109/TASE.2015. 2412256

    work page doi:10.1109/tase.2015 2016

  21. [21]

    In: Advances in Service and Industrial Robotics

    Žlajpah, L., Petrič, T.: Virtual guides for redundant robots using admittance con- trol for path tracking tasks. In: Advances in Service and Industrial Robotics. pp. 13–23 (2019).https://doi.org/10.1007/978-3-030-00232-9_2

    work page doi:10.1007/978-3-030-00232-9_2 2019

  22. [22]

    Zolotas, M., Wonsick, M., Long, P., Padır, T.: Motion polytopes in virtual reality for shared control in remote manipulation applications. Front. Robot. AI (2021). https://doi.org/10.3389/frobt.2021.730433

    work page doi:10.3389/frobt.2021.730433 2021