Pith. sign in

REVIEW 4 major objections 4 minor 36 references

Predictive Relative-Velocity Steering for Safe Robotic Manipulator Teleoperation in Dynamic Environments

T0 review · 4 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

Pith's one-line read PARRY redirects the relative-velocity vector instead of adding repulsive forces, and that change lowers teleoperated collision rates by up to 38.7%.

desk verdict Reasonable incremental safety layer for teleoperation with a solid paired Monte Carlo evaluation, but the latency-compensation story is overstated and the hand-tuned parameters need sensitivity analysis. read the letter →

arxiv 2608.13284 v1 pith:HX27F3QU submitted 2026-08-13 cs.RO

classification cs.RO
keywords teleoperationcollisionavoidancerelative-velocityrotationtime-to-collisionpredictionpoint-cloudpreprocessingartificialpotentialfieldcontrolbarrierfunctionobstacle-observationlatency
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

This paper argues that a teleoperated robot arm can avoid suddenly appearing obstacles more reliably if a safety layer redirects the operator's velocity command in the obstacle's reference frame rather than superimposing repulsive forces. The proposed method, PARRY, predicts where the point cloud will be on a time-to-collision horizon, then rotates the relative-velocity vector by a repulsion-dependent angle while keeping its magnitude unchanged. In 1,000 paired Monte Carlo trials per condition, this lowers the end-effector collision rate by 17.2%, 27.3%, and 38.7% relative to three baselines, and the advantage persists when obstacle observations are delayed by 100 or 150 ms. A physical 7-degree-of-freedom setup shows the same steering behavior under static and dynamic obstacles. The result matters because it targets two practical failure modes of teleoperation safety: deadlock-like velocity stagnation and perception latency.

What carries the argument

The central object is the relative-velocity vector $v_{\mathrm{rel}} = v_{\mathrm{ee}} - v_{\mathrm{o}}$ together with the Rodrigues rotation formula $v_s = v_{\mathrm{rel}} \cos\theta + (\hat{a} \times v_{\mathrm{rel}}) \sin\theta + \hat{a}(\hat{a}^T v_{\mathrm{rel}})(1-\cos\theta)$. The rotation axis $\hat{a}$ is perpendicular to both the relative velocity and the aggregate repulsive force, and the angle $\theta$ is clipped to a maximum value. This carries the argument by converting collision avoidance into a pure redirection of the operator's intent, preserving speed and avoiding deadlock. Around it sits the prediction module: the most hazardous point's velocity is estimated by frame differencing with exponential moving average smoothing, a time-to-collision-based horizon $t_p$ adapts to radial approach speed, and overshoot protection clamps predicted points so they do not pass beyond the end effector and create reverse repulsive forces.

What would settle it

Run the same 1,000-trial paired protocol with obstacles that accelerate by more than 1 m/s within the 100-150 ms observation window and compare PARRY at 150 ms latency against its no-prediction ablation. If the collision rate falls to the ablation level, the constant-velocity nearest-point velocity estimate is the load-bearing component.

Watch

Extended reading notes

Core claim

The paper's central claim is that PARRY makes reactive collision avoidance for teleoperated manipulators safer by steering the relative-velocity vector $v_{\mathrm{rel}} = v_{\mathrm{ee}} - v_{\mathrm{o}}$ rather than adding repulsive forces to the task velocity. It computes a repulsive direction from a predicted point cloud, selects a rotation axis perpendicular to both relative velocity and repulsive force, and applies Rodrigues' rotation formula to turn $v_{\mathrm{rel}}$ by an angle proportional to repulsion magnitude, bounded by a maximum. Because only the direction changes, the commanded speed never collapses toward zero, so the operator does not experience the velocity stagnation typical of APF local minima. Around this sits a prediction module that reconstructs a look-ahead point cloud on a time-to-collision horizon, with overshoot protection that prevents predicted points from passing beyond the end effector. The paper reports that PARRY keeps the lowest collision rate at 0, 100, and 150 ms obstacle-observation latency while adding only a median of 0.101 ms per cycle.

Load-bearing premise

The latency-compensation scheme assumes that the nearest obstacle point's velocity stays constant over the short prediction horizon; if that point switches between frames or the obstacle accelerates, the predicted point cloud is wrong and the claimed collision-avoidance advantage breaks down.

Editorial extensions

If this is right

  • At 0, 100, and 150 ms obstacle-observation latency, PARRY keeps the lowest collision avoidance rate among APF-VS, CBF-QP, SSM, and its own ablations.
  • Relative-velocity deflection reduces the end-effector collision rate by 17.2%, 27.3%, and 38.7% relative to APF-VS, CBF-QP, and SSM at baseline latency.
  • TTC-based prediction limits the collision-rate degradation from 0 to 150 ms latency to 2.6 percentage points, compared with 5.4 points for APF-VS and 4.3 points for CBF-QP.
  • Overshoot protection cuts the fraction of trials with large-angle repulsive-force flips from 20.2% to 3.6% and wrong-way repulsion from 20.4% to 0.1% at 0 ms latency.
  • In symmetric two-sphere configurations where APF methods stall, PARRY preserves a forward velocity component and reaches the target earlier.

Reading between the lines

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

  • An implication the paper leaves implicit is that PARRY's constant-velocity, nearest-point obstacle model will degrade sharply for accelerating or articulating obstacles; replacing the single-point estimate with per-cluster velocity estimates would be a direct and testable extension.
  • Because the rotation preserves relative-velocity magnitude, PARRY behaves like a mirror-reflection law in the obstacle frame; a testable hypothesis is that repeated applications drive the end effector onto a trajectory tangent to the obstacle surface, which could be compared analytically with classical APF equilibria.
  • Since the overshoot guard mostly improves motion smoothness rather than collision rate, its largest practical benefit may be operator trust and reduced false alarms under real human-in-the-loop teleoperation, which the fixed-trajectory Monte Carlo setup does not measure.
  • PARRY's latency-compensated point-cloud prediction is a general preprocessing step: the same TTC-based look-ahead and overshoot guard could be attached to other velocity-level safety filters, such as CBF-based projections, to give them the same delay robustness.
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

4 major / 4 minor

Summary. The paper proposes PARRY, a reactive collision-avoidance framework for teleoperated manipulators that operates at the end-effector velocity-command level. It preprocesses a depth point cloud, estimates the obstacle velocity from the nearest point, generates a TTC-based predicted point cloud with overshoot protection, and deflects the relative velocity vector via Rodrigues' rotation while preserving its magnitude. The authors evaluate PARRY against APF-VS, CBF-QP, and SSM in paired Monte Carlo simulations in MuJoCo (1,000 trials per method per latency condition at 0, 100, and 150 ms), report higher collision-avoidance rates, ablations isolating prediction, overshoot protection, and relative-velocity processing, and demonstrate the approach on a physical 7-DoF Flexiv Rizon 4 platform.

Significance. If the results hold, PARRY is a practical and lightweight safety layer for teleoperation: it directly processes point clouds, runs at 0.1 ms per cycle, and avoids the deadlock issue of classic APF. The paired Monte Carlo design with exact McNemar tests and Holm correction is appropriate, and the ablations give useful evidence that TTC prediction and relative-velocity redirection each contribute. The main limitations are lack of a defined collision criterion, undocumented parameter calibration, and the unsupported mechanistic claim that prediction compensates specifically for latency. These are fixable with added details and experiments; the central empirical comparison is not obviously invalid.

major comments (4)
  1. [Section IV-A and Table I] The primary metric CAR (collision avoidance rate) is never defined. The manuscript must specify the exact collision event used in MuJoCo (e.g., distance between the end-effector sphere and obstacle sphere falling below the sum of radii, or contact reported by the physics engine) and the time window/sampling rule over which collisions are counted. Without this definition, the central numerical comparisons cannot be reproduced or interpreted.
  2. [Section IV-B (Baselines and Evaluation Metrics)] The statement that 'the key parameters of each method are calibrated independently' is not accompanied by any parameter values or description of the calibration protocol. PARRY alone has more than ten tunable constants (k_rot, θ_max, d_max, p, β, t_min, t_max, α, d_j, d_s, d_p, and the ε thresholds). To support the claim that PARRY outperforms baselines rather than being better tuned, please provide a table of all parameters and describe how they were selected, including whether calibration was performed on a held-out set.
  3. [Section III-C, Eq. (6), and Section IV-C.2] The prediction horizon t_p is defined as a clipped function of the radial approach time T_app, with no term proportional to the injected obstacle-observation latency τ. Since the same TTC schedule is used for the 0, 100, and 150 ms conditions, the paper's claim that the prediction module 'compensates for obstacle-observation latency' is not mechanically supported; the observed latency robustness could equally arise from a generic forward shift of the point cloud. Please either tie t_p explicitly to τ (e.g., t_p = βT_app + τ), or reframe the contribution as 'TTC-based look-ahead that is robust to latency' and add an experiment that varies τ while holding the prediction horizon fixed to isolate the compensation effect.
  4. [Section III-C, Eqs. (4)-(5)] The obstacle velocity estimate v_o is obtained by differencing the single most hazardous point in consecutive frames without data association, so the estimate is correct only if the nearest point corresponds to the same physical surface point across frames. The EMA and jump threshold d_j mitigate but do not solve nearest-point switching. Since Eq. (7) shifts the entire point cloud by this single velocity, any estimation error propagates to all predicted points. The paper should report the velocity-estimation error in the simulated scenarios and test a multi-obstacle or non-smooth case in which nearest-point switching is likely.
minor comments (4)
  1. [Figure 1 caption] The caption contains a stray 'Prompt: A 2x2 scientific illustration...' text; remove it.
  2. [Section IV-C.3] The sentence 'After TTC-based point-cloud prediction is introduced, PARRY w/o Guard reduces the collision rate by 17.5%, 21.7%, and 24.9% relative to PARRY w/o Pred' is correct only if the reader understands that the comparison is between collision rates (1-CAR), not CAR values; please state this explicitly in the text.
  3. [Section III-D, Algorithm 1] The initial rotation axis \hat{a}^{(0)} is only constrained to be perpendicular to v_ee; its arbitrary choice may bias the first redirection. Report or specify a deterministic initialization.
  4. [Section V] The physical-robot experiments are purely qualitative. Please report at least the number of trials and success/failure counts, or describe them explicitly as demonstrative rather than as validation.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the PARRY derivation is self-contained and its claims are not reduced to their inputs by construction.

full rationale

Walking the derivation chain, no step exhibits the quoted-reduction form that circularity requires. The commanded velocity v_cmd is obtained as v_s + v_o, where v_s is a Rodrigues rotation of v_rel = v_ee - v_o; the rotation axis and angle are computed from the predicted point cloud and from F_rep, which is itself a distance-weighted function of the predicted obstacle points. None of these quantities is defined in terms of v_cmd or in terms of the collision-avoidance-rate metric that is later reported. The prediction equations (4)-(9) take the current point cloud, a nearest-point velocity estimate, and an EMA-filtered v_o as inputs, and produce a shifted, overshoot-protected cloud. This is a standard model-based forward shift rather than a fitted rearrangement of the output, so it is not a self-definitional reduction. The ablation and baseline comparisons in Table I are empirical measurements, not derivations from the method's own assumptions, so the central quantitative claim is independently grounded in the Monte Carlo trials. The self-citations in the paper ([3], [10], [26], [28], [30], [33]) appear only as background, related-work context, and supplementary safety-layer references; they are not used to justify the uniqueness or correctness of PARRY's construction, and no load-bearing argument is replaced by a self-citation. Legitimate concerns about the nearest-point velocity estimate (Eq. 4-6) and about whether the TTC-based prediction horizon, which does not contain the measured latency tau, truly compensates for the injected 100/150 ms delay are correctness and validity risks, not circularity: the equations still take independent inputs and produce a well-defined output, and the paper does not define its prediction metric in terms of its own results. Therefore the honest finding is no significant circularity.

Assumptions & free parameters 10 free parameters · 5 assumptions · 0 invented entities

PARRY introduces no new physical entities. It relies on several domain assumptions about obstacle motion and point-cloud fidelity, and on a set of hand-tuned hyperparameters whose values and calibration protocol are not reported. The method's predictive component is a constant-velocity geometric extrapolation, not a learned or fitted model.

free parameters (10)
  • k_rot = not reported
    Gain mapping repulsive force magnitude to deflection angle (Eq. 15); hand-tuned.
  • theta_max = not reported
    Maximum allowed deflection angle (Eq. 15); hand-tuned.
  • d_max = not reported
    Obstacle influence distance for active set and repulsive force (Eq. 10-13); hand-tuned.
  • p = not reported
    Distance-decay hyperparameter in power-law repulsive force (Eq. 12); p > 1, hand-tuned.
  • beta = not reported
    Scaling factor for TTC-based prediction horizon (Eq. 6); hand-tuned.
  • t_min, t_max = not reported
    Bounds on prediction horizon (Eq. 6); hand-tuned.
  • alpha = not reported
    EMA smoothing factor for obstacle velocity (Eq. 5); hand-tuned.
  • d_j = not reported
    Displacement threshold that resets velocity estimate to previous frame (Eq. 5); hand-tuned.
  • d_s, d_p = not reported
    Minimum predicted-point distance and look-ahead cap in overshoot protection (Eq. 8 and Sec. III-C); hand-tuned.
  • epsilon thresholds (epsilon_a, epsilon_o, epsilon_f, epsilon_v, epsilon_x) = not reported
    Small positive thresholds for numerical stability and degeneracy handling (Sec. III-A and Eq. 14); hand-tuned.
assumptions (5)
  • standard math Rodrigues' rotation formula correctly rotates a 3D vector about a unit axis by a given angle (Eq. 16).
    Standard linear algebra result, used without proof.
  • domain assumption Obstacle velocity is constant over the prediction horizon (Eq. 7).
    The predicted point cloud translates each point by v_o * t_p, assuming uniform motion.
  • domain assumption The most hazardous point's velocity represents the whole obstacle's motion (Eq. 4-6).
    All points in the cloud are shifted by the velocity estimated from the single nearest point.
  • domain assumption The preprocessed point cloud faithfully represents obstacle geometry (Sec. III-B).
    Voxel downsampling, self-filtering, and outlier removal may discard or distort points needed for avoidance.
  • domain assumption End-effector-level collision avoidance is a sufficient safety objective for the tested teleoperation tasks (Sec. VI).
    The paper explicitly limits claims to end-effector collisions; whole-arm collisions are out of scope.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Predictive Relative-Velocity Steering for Safe Robotic Manipulator Teleoperation in Dynamic Environments." pith.science (2026). https://pith.science/paper/HX27F3QU

@misc{pith2026260813284,
  author       = {Pith},
  title        = {Pith review of: Predictive Relative-Velocity Steering for Safe Robotic Manipulator Teleoperation in Dynamic Environments},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HX27F3QU}},
  note         = {Machine review of arXiv:2608.13284}
}
read the original abstract

Recent advances in teleoperation have enabled robotic manipulators to perform dexterous, human-arm-like motions. However, human operators may fail to avoid suddenly appearing obstacles promptly and effectively, particularly under network latency or limited attention, thereby creating safety risks. To address this issue, we propose a lightweight and modular framework for proactive collision avoidance, operating directly at the end-effector velocity-command level. After preprocessing the point cloud, the framework first predicts potential collisions based on time-to-collision (TTC) with integrated overshoot protection, and subsequently rotates the relative-velocity vector using Rodrigues' rotation formula. The deflection changes only the direction of the relative velocity while preserving its magnitude, thereby mitigating the deadlock problem commonly encountered by conventional artificial potential field (APF) methods. The prediction module compensates for point-cloud processing latency introduced by complex teleoperation pipelines, while the lightweight design enables the high-frequency control required for teleoperation. Simulations across diverse scenarios show that the proposed method achieves a higher end-effector collision avoidance rate than the baseline methods. Experiments on a physical robotic system further validate its collision-avoidance effectiveness.

Figures

Figures reproduced from arXiv: 2608.13284 by the authors.

Figure 1
Figure 1. Illustration of PARRY in teleoperation. Prompt: A 2x2 scientific illustration. Top row: Male VR operator faces network lag, causing a robotic arm to crash into a grey pillar (red explosion). Bottom row: Female VR operator works smoothly, robotic arm uses overhead camera vision to deflect its path along a green curve, successfully avoiding the pillar. teleoperation from several directions. These mechanisms include re… view at source ↗
Figure 2
Figure 2. Overall architecture of PARRY. Algorithm 1: PARRY Velocity Deflection Input: Point cloud X (k) = {x b i}; end-effector state (x (k) ee , v (k) ee ); previous state S (k−1) Output: v (k) cmd and S (k) 1 Init.: Set v (0) o ← 0 and choose aˆ (0) ⊥ v (0) ee ; 2 Obtain O (k) by preprocessing X (k) ; 3 if O (k) = ∅ then 4 return v (k) ee , S (k−1); 5 else 6 i ∗ ← arg min i ∥o (k) i − x (k) ee ∥2; 7 Estimate v (k) o from o… view at source ↗
Figure 3
Figure 3. Comparison between APF velocity superposition and PARRY [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: In the MuJoCo simulation, PARRY redirects the end effector around [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Monte Carlo evaluation (a) CAR under increasing obstacle [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: End-effector forward velocity vy and position y for APF-VS (dashed) and PARRY (solid) under four symmetric configurations. which APF methods are prone to local minima. As shown in [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: Static obstacle scenario. Top: Baseline collision. Bottom: PARRY successfully steers around the obstacle in real time [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]
Figure 8
Figure 8. Figure 8: Physical-robot trajectories. B. Evaluation in Static and Dynamic Scenarios We evaluate PARRY under both static blockage and dy￾namic intrusions. In the static scenario ( [PITH_FULL_IMAGE:figures/full_fig_p007_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

36 extracted references · 26 canonical work pages

  1. [1]

    Bilateral teleoperation: An histor- ical survey,

    P. F. Hokayem and M. W. Spong, “Bilateral teleoperation: An histor- ical survey,”Automatica, vol. 42, no. 12, pp. 2035–2057, 2006

  2. [2]

    Attentiveness map estimation for haptic teleoperation of mobile robot obstacle avoidance and approach,

    N. Zhong and K. Hauser, “Attentiveness map estimation for haptic teleoperation of mobile robot obstacle avoidance and approach,”IEEE Robotics and Automation Letters, vol. 9, no. 3, pp. 2152–2159, 2024

  3. [3]

    Real-time safety assessment techniques of dynamic systems,

    X. He, Z. Liu, S. Hu, C. Liu, and D. Zhou, “Real-time safety assessment techniques of dynamic systems,”Acta Automatica Sinica, vol. 51, no. 2, pp. 249–270, 2024

  4. [4]

    Shared control in robot teleoperation with improved potential fields,

    A. Gottardi, S. Tortora, E. Tosello, and E. Menegatti, “Shared control in robot teleoperation with improved potential fields,”IEEE Transac- tions on Human-Machine Systems, vol. 52, no. 3, pp. 410–422, 2022

  5. [5]

    Real-time obstacle avoidance for manipulators and mobile robots,

    O. Khatib, “Real-time obstacle avoidance for manipulators and mobile robots,”The International Journal of Robotics Research, vol. 5, no. 1, pp. 90–98, 1986

  6. [6]

    A shared control method for teleoperated robot using artificial potential field,

    W. Qin, H. Zhang, Z. Fan, Y . Zhu, and J. Zhao, “A shared control method for teleoperated robot using artificial potential field,”Advanced Intelligent Systems, vol. 6, no. 10, p. 2300814, 2024

  7. [7]

    Virtual fixtures: Perceptual tools for telerobotic manipulation,

    L. B. Rosenberg, “Virtual fixtures: Perceptual tools for telerobotic manipulation,” inProceedings of IEEE Virtual Reality Annual Inter- national Symposium. IEEE, 1993, pp. 76–82

  8. [8]

    Method for generating real-time interactive virtual fixture for shared teleoperation in unknown environments,

    V . Pruks and J.-H. Ryu, “Method for generating real-time interactive virtual fixture for shared teleoperation in unknown environments,”The International Journal of Robotics Research, vol. 41, no. 9-10, pp. 925– 951, 2022

Show all 36 references
  1. [9]

    Control-Barrier-Aided Teleoperation with Visual-Inertial SLAM for Safe MA V Navigation in Complex Environments,

    S. Zhou, S. Papatheodorou, S. Leutenegger, and A. P. Schoellig, “Control-Barrier-Aided Teleoperation with Visual-Inertial SLAM for Safe MA V Navigation in Complex Environments,” in2024 IEEE International Conference on Robotics and Automation (ICRA), 2024, pp. 17 836–17 842

  2. [10]

    Safelink: Safety-critical control under dynamic and irregular unsafe regions,

    S. Hu, Z. Wang, Z. Liu, Z. Shen, and X. He, “Safelink: Safety-critical control under dynamic and irregular unsafe regions,” 2026. [Online]. Available: https://arxiv.org/abs/2503.16551

  3. [11]

    Implementing speed and separation monitoring in collaborative robot workcells,

    J. A. Marvel and R. Norcross, “Implementing speed and separation monitoring in collaborative robot workcells,”Robotics and Computer- Integrated Manufacturing, vol. 44, pp. 144–155, 2017

  4. [12]

    Potential field methods and their inherent limitations for mobile robot navigation,

    Y . Koren and J. Borenstein, “Potential field methods and their inherent limitations for mobile robot navigation,” inProceedings. 1991 IEEE International Conference on Robotics and Automation, 1991, pp. 1398–1404 vol.2

  5. [13]

    Real-time robot motion control with circulatory fields,

    L. Singh, H. Stephanou, and J. Wen, “Real-time robot motion control with circulatory fields,” inProceedings of IEEE International Confer- ence on Robotics and Automation, vol. 3. IEEE, 1996, pp. 2737–2742

  6. [14]

    Circular fields and predictive multi-agents for online global trajectory planning,

    M. Becker, T. Lilge, M. A. M ¨uller, and S. Haddadin, “Circular fields and predictive multi-agents for online global trajectory planning,” IEEE Robot. Autom. Lett., vol. 6, no. 2, pp. 2618–2625, 2021

  7. [15]

    Informed circular fields for global reactive obstacle avoidance of robotic manipulators,

    M. Becker, P. Caspers, T. Hattendorf, T. Lilge, S. Haddadin, and M. A. M¨uller, “Informed circular fields for global reactive obstacle avoidance of robotic manipulators,”IFAC-PapersOnLine, vol. 56, no. 2, pp. 1017–1022, 2023

  8. [16]

    Magnetic-field-inspired nav- igation for robots in complex and unknown environments,

    A. Ataka, H.-K. Lam, and K. Althoefer, “Magnetic-field-inspired nav- igation for robots in complex and unknown environments,”Frontiers in Robotics and AI, vol. 9, p. 834177, 2022

  9. [17]

    An optimization-based magnetic field potential learning framework for robotic volumetric obstacle-avoidance,

    S. Yang, Y . Sun, X. Su, J. Huang, Z. Bing, and A. Knoll, “An optimization-based magnetic field potential learning framework for robotic volumetric obstacle-avoidance,”IEEE Transactions on Indus- trial Electronics, 2025

  10. [18]

    Modified artificial potential field for the path planning of aircraft swarms in three-dimensional environments,

    R. M. J. A. Souza, G. V . Lima, A. S. Morais, L. C. Oliveira- Lopes, D. C. Ramos, and F. L. Tofoli, “Modified artificial potential field for the path planning of aircraft swarms in three-dimensional environments,”Sensors, vol. 22, no. 4, p. 1558, 2022

  11. [19]

    A dynamical system approach to realtime obstacle avoidance,

    S. M. Khansari-Zadeh and A. Billard, “A dynamical system approach to realtime obstacle avoidance,”Autonomous Robots, vol. 32, no. 4, pp. 433–454, 2012

  12. [20]

    Avoidance of concave obsta- cles through rotation of nonlinear dynamics,

    L. Huber, J.-J. Slotine, and A. Billard, “Avoidance of concave obsta- cles through rotation of nonlinear dynamics,”IEEE Transactions on Robotics, vol. 40, pp. 1983–2002, 2023

  13. [21]

    Care: Enhancing safety of visual navigation through collision avoidance via repulsive estimation,

    J. Kim, J. Sim, W. Kim, K. Sycara, and C. Nam, “Care: Enhancing safety of visual navigation through collision avoidance via repulsive estimation,”arXiv preprint arXiv:2506.03834, 2025

  14. [22]

    Escaping local minima: Hybrid artificial potential field with wall- follower for decentralized multi-robot navigation,

    J. Kim, S. Park, W. Lee, W. Kim, H. Choi, N. Doh, and C. Nam, “Escaping local minima: Hybrid artificial potential field with wall- follower for decentralized multi-robot navigation,” in2025 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2025, pp. 6616–6622

  15. [23]

    Smith predictor type control ar- chitectures for time delayed teleoperation,

    A. C. Smith and K. Hashtrudi-Zaad, “Smith predictor type control ar- chitectures for time delayed teleoperation,”The International Journal of Robotics Research, vol. 25, no. 8, pp. 797–818, 2006

  16. [24]

    Model-mediated telemanipulation,

    P. Mitra and G. Niemeyer, “Model-mediated telemanipulation,”The International Journal of Robotics Research, vol. 27, no. 2, pp. 253– 262, 2008

  17. [25]

    Model-mediated delay compensation with goal prediction for robot teleoperation over internet,

    R. Lima, V . Vakharia, U. Rai, H. Mehta, V . Vatsal, and K. Das, “Model-mediated delay compensation with goal prediction for robot teleoperation over internet,” in2023 IEEE International Conference on Systems, Man, and Cybernetics (SMC). IEEE, 2023, pp. 1037–1043

  18. [26]

    Real-time safety assessment of dynamic systems in non-stationary environments: A review of methods and techniques,

    Z. Liu, S. Hu, and X. He, “Real-time safety assessment of dynamic systems in non-stationary environments: A review of methods and techniques,” in2023 CAA Symposium on Fault Detection, Supervision and Safety for Technical Processes (SAFEPROCESS). IEEE, 2023, pp. 1–6

  19. [27]

    Understanding and mitigating network latency effect on teleoperated-robot with extended reality,

    Z. Zhang, C. Liu, and H. Kim, “Understanding and mitigating network latency effect on teleoperated-robot with extended reality,”arXiv preprint arXiv:2506.01135, 2025

  20. [28]

    CADM+: Confusion-based learning framework with drift detection and adaptation for real-time safety assessment,

    S. Hu, Z. Liu, M. Li, and X. He, “CADM+: Confusion-based learning framework with drift detection and adaptation for real-time safety assessment,”IEEE Transactions on Neural Networks and Learning Systems, vol. 36, no. 3, pp. 5126–5139, 2024

  21. [29]

    Time-to-contact for robot safety stop in close collaborative tasks,

    A. Olivares-Alarcos, S. Foix, and G. Aleny `a, “Time-to-contact for robot safety stop in close collaborative tasks,” inHuman-Robot Col- laboration. IET, 2023, ch. Chapter 5, pp. 87–104

  22. [30]

    Online dynamic hybrid broad learning system for real-time safety assessment of dynamic systems,

    Z. Liu and X. He, “Online dynamic hybrid broad learning system for real-time safety assessment of dynamic systems,”IEEE Transactions on Knowledge and Data Engineering, vol. 36, no. 12, pp. 8928–8938, 2024

  23. [31]

    MuJoCo: A physics engine for model-based control,

    E. Todorov, T. Erez, and Y . Tassa, “MuJoCo: A physics engine for model-based control,” in2012 IEEE/RSJ International Conference on Intelligent Robots and Systems. IEEE, 2012, pp. 5026–5033

  24. [32]

    Control barrier function based quadratic programs for safety critical systems,

    A. D. Ames, X. Xu, J. W. Grizzle, and P. Tabuada, “Control barrier function based quadratic programs for safety critical systems,”IEEE Trans. Autom. Control, vol. 62, no. 8, pp. 3861–3876, 2016

  25. [33]

    VLSA: Vision-language-action models with plug- and-play safety constraint layer,

    S. Hu, Z. Liu, S. Liu, J. Cen, Z. Meng, S. Wang, X. Li, and X. He, “VLSA: Vision-language-action models with plug- and-play safety constraint layer,” 2026. [Online]. Available: https://arxiv.org/abs/2512.11891

  26. [34]

    Note on the sampling error of the difference between correlated proportions or percentages,

    Q. McNemar, “Note on the sampling error of the difference between correlated proportions or percentages,”Psychometrika, vol. 12, no. 2, pp. 153–157, 1947

  27. [35]

    A simple sequentially rejective multiple test procedure,

    S. Holm, “A simple sequentially rejective multiple test procedure,” Scandinavian Journal of Statistics, pp. 65–70, 1979

  28. [36]

    Reactive diffusion policy: Slow-fast visual-tactile policy learning for contact-rich manipulation,

    H. Xue, J. Ren, W. Chen, G. Zhang, Y . Fang, G. Gu, H. Xu, and C. Lu, “Reactive diffusion policy: Slow-fast visual-tactile policy learning for contact-rich manipulation,”arXiv preprint arXiv:2503.02881, 2025

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.