Pith. sign in

REVIEW 4 major objections 6 minor 27 references

Interactive OT Gym: A Reinforcement Learning-Based Interactive Optical tweezer (OT)-Driven Microrobotics Simulation Platform

T0 review · 4 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A shared human-RL controller improves optical-tweezer micromanipulation in simulation, cutting task time by roughly two-thirds and reaching 100% success.

desk verdict A genuinely useful simulation platform whose headline performance claim overstates what the data show; worth refereeing after a honest rewrite. read the letter →

arxiv 2505.20751 v2 pith:2XP226NU submitted 2025-05-27 cs.RO cs.AI

Zongcai Tan , Dandan Zhang This is my paper
classification cs.ROcs.AI
keywords opticaltweezersmicroroboticsreinforcementlearningsharedcontrolhapticfeedbacksimulationplatformmicromanipulationpathplanning
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

Interactive OT Gym is a simulation platform for optical-tweezer-driven microrobots that combines a physics-based microfluidic simulator, haptic teleoperation, and reinforcement-learning autonomous control. The paper's central claim is that a context-aware shared-control strategy, which blends human commands with RL navigation according to distance to obstacles, substantially improves cooperative cell-transport performance. In the reported experiments, shared control completed the task in 53.9 seconds with 100% success, versus 163.5 seconds and 94% success for manual control and 54.1 seconds and 59% success for autonomous RL control alone. If correct, the result matters because it offers an inexpensive, interactive training and testing ground for human-robot collaborative micromanipulation before committing to physical experiments.

What carries the argument

The load-bearing mechanism is the context-aware shared-control law of Eq. (5), which sets the human weight alpha to 0.5 when the robot is close to an obstacle (d <= d1), fades it to 0.1 when the obstacle is far (d > d2), and interpolates in between, so the operator is pulled in for fine-tuning exactly when the environment gets risky. Around this, the platform combines a progressive RL navigation scheme (A* with B-spline smoothing for global paths, and a DQN for local speed control), a bimanual haptic interface driven by the empirical trap-force model of Eq. (4), and a distributed ROS-based architecture. This stack is what lets the human and the RL agent exchange authority seamlessly during a task.

What would settle it

Run the same cell-transport task on a physical multi-trap optical tweezer with haptic teleoperation and the same shared-control law; if shared control does not beat autonomous RL in success rate, or if the scaled haptic forces of Eq. (4) do not reproduce measured trap forces, the platform's central claim fails.

Watch

Extended reading notes

Core claim

The paper introduces what it describes as the first interactive OT microrobotics platform to combine high-fidelity physical-field simulation, haptic feedback, RL-based autonomous navigation, and shared control. The central discovery is that mixing human teleoperation with a DQN-based speed controller through a distance-dependent blending weight yields better overall task performance than either mode alone: in the cell-transport task, shared control reached a 100% success rate with roughly two-thirds less time than manual control, while pure RL control lost cells against dynamic obstacles and managed only 59% success. The reported completion times show the speed advantage is almost entirely over manual control, and the success-rate improvement over RL is what the human-in-the-loop contribution buys.

Load-bearing premise

The load-bearing premise is that the simulated microfluidic and optical-tweezer physics, especially the empirically scaled force model that drives haptic feedback, matches real micro-scale behavior closely enough that performance gains measured in the simulator will carry over to physical systems.

Editorial extensions

If this is right

  • Shared control with context-aware weighting can let a single operator outperform both full teleoperation and full autonomy in dynamic, obstacle-rich microenvironments.
  • RL-based dynamic speed control maintains 100% success while reducing completion time relative to any fixed speed, and it avoids the trap-loss failures of fast constant speeds.
  • B-spline smoothing of A* paths produces trajectories with lower curvature, lower angular deviation, and less high-frequency jitter, which suits precision micromanipulation.
  • The platform provides a low-cost environment for training RL agents and for testing interactive control algorithms before physical experiments are run.

Reading between the lines

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

  • Looking at the paper's own tables, the 67% time saving is measured against manual control; autonomous RL already took 54.1 seconds versus 53.9 seconds for shared control, so the real added value of the human-RL blend here is reliability (100% versus 59% success), not speed.
  • The distance-based blending rule of Eq. (5) is a generic mechanism that could be lifted into other human-robot micromanipulation domains, such as magnetic or optoelectronic microrobots, whenever a distance-to-obstacle signal is available.
  • A physical experiment with the same microrobot and shared-control law would be the natural next test; until then, the reported result is a simulation-platform claim rather than a hardware demonstration.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 6 minor

Summary. The paper presents Interactive OT Gym, a simulation platform for optical-tweezer-driven microrobots built on NVIDIA Isaac Sim, integrating a DQN-based speed controller over A* paths, bimanual haptic teleoperation, and a context-aware shared-control mode that blends human and autonomous commands. The authors evaluate path smoothness, RL training convergence, RL dynamic speed control, and a user study with eight volunteers comparing manual, autonomous, and shared control on a simulated cooperative cell-transport task. The headline claim is that shared control reduces task completion time by approximately 67% compared with either pure human or RL control, with a 100% success rate, while also lowering NASA-TLX workload and improving UEQ-S scores.

Significance. If the platform and results were fully supported, the paper would make a useful contribution: a low-cost, interactive simulation environment for OT-driven microrobotics that combines haptic feedback, RL-based autonomy, and shared control is a plausible training and development tool for the robotics and micromanipulation communities. The progressive A* plus DQN architecture is a sensible division of labor, and the human-subject evaluation with NASA-TLX and UEQ-S is a positive feature that goes beyond pure simulation metrics. However, the central quantitative claim is not supported by the reported data, the success-rate metric is defined inconsistently across experiments, and the claimed physical fidelity is not validated. These issues materially weaken the paper's main contributions as currently stated, though they are addressable with revised claims and additional validation.

major comments (4)
  1. [Abstract; Section III-D; Fig. 7(a)] The Abstract and Conclusion claim a ~67% reduction in task completion time 'compared to using pure human or RL control alone'. Figure 7(a) reports manual 163.5±8.3 s, autonomous 54.1±0.1 s, and shared 53.9±0.2 s. The shared-control time is only about 0.4% shorter than autonomous control, and no statistical test is reported for this pairwise comparison. The 67% figure is valid only against manual control. Please revise the claim to name the comparison baseline explicitly and add a statistical comparison for shared versus autonomous control.
  2. [Section III-C vs. Section III-D; Abstract] The success rate is defined twice: in Section III-C as 'the percentage of the total distance completed' and in Section III-D as 'the proportion of dynamical obstacles avoided'. The Abstract's '100% success rate' is therefore ambiguous. Please define a single primary success metric, report it consistently across all control modes, and state explicitly which metric the 100% figure refers to.
  3. [Table II vs. Fig. 7(b)] Section III-C reports that the RL-based controller achieves a 100% average success rate (Table II), yet the autonomous condition in Section III-D is reported at 59±5% success (Fig. 7(b)). Since the shared-control comparison in Section III-D uses the autonomous controller as its baseline, the two evaluations must differ in task difficulty, environment settings, or controller configuration. Please explain this discrepancy and clarify which task setup applies to the shared-control comparison.
  4. [Sections II-B and II-D, Eq. (4)] The platform is described as high-fidelity and Eq. (1) includes optical, Brownian, van der Waals, drag, hydrodynamic, and contact forces, but the haptic feedback force model in Eq. (4) uses parameters K=0.455, δ=0.446, A=0.058, C=0.01 that are only described as 'empirically scaled' in Section II-D. No calibration data, physical reference measurements, or comparison with real OT experiments are reported. The 'high fidelity' claim and the validity of the haptic feedback experience are not established; please add calibration or validation evidence, or temper the fidelity claim accordingly.
minor comments (6)
  1. [Section III-B] The phrase 'navigation navigation efficiency' contains a duplicated word; remove the repetition.
  2. [Section II-C] The training description contains the stray text 'colorredwith 8GB', apparently a leftover LaTeX color command; remove it.
  3. [Section III-D] The text reads 'ANOV A' instead of 'ANOVA'; fix the spacing.
  4. [Table II] The caption does not explain the group labels G1 through G6 or the slash notation, which are defined only in the body text; add a concise note in the caption for readability.
  5. [Section III-D, Experimental Setup] Please clarify whether each of the eight trials per mode involved a distinct human run and whether autonomous and shared runs were performed as human-in-the-loop trials or as repeated simulation episodes; this affects the interpretation of the reported mean and error bars.
  6. [Section II-E, Eq. (5)] The parameters d1 and d2 in Eq. (5) are stated as d1=1, d2=2 in Fig. 4 but without units; specify the units and whether these values are measured in the simulation's physical coordinates.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the shared-control result is an empirical simulation measurement, not a consequence of the fitted or hand-tuned parameters.

full rationale

The paper's derivation chain is not circular. Equations (1)-(5) define a physics simulator, an RL reward, a haptic force model, and a shared-control blending rule; none of these is used to derive the outcome claims. The claimed 67% time reduction and 100% success rate are presented as measured results from simulated trials (Fig. 7, Tables I-II), not as predictions entailed by the model equations. The force-model parameters in Eq. (4) are 'empirically scaled' and the α thresholds in Eq. (5) are hand-set, but the comparison of manual, autonomous, and shared control is an experimental evaluation in the simulator, so those values are inputs to the controller, not to the evaluation outcome. The one self-citation to Zhang et al. [13] only supplies the microrobot geometry used in the simulation and is not load-bearing for any claimed result. The abstract's '67%' statement is internally inconsistent with Fig. 7(a) because shared control (53.9 s) is essentially equal to autonomous control (54.1 s) and the 67% reduction holds only against manual control (163.5 s); however, this is a correctness/consistency defect, not a circularity. Likewise, the redefinition of 'success rate' between Sections III-C and III-D makes the abstract's 100% claim ambiguous, but it does not reduce any derived result to its inputs. No circular step can be quoted and exhibited, so the appropriate circularity score is 0.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The central results depend on a chain of modeling assumptions and empirically tuned parameters. No new physical entities are introduced. The free parameters are the trap force model constants, shared-control distance thresholds, and reward shaping thresholds.

free parameters (6)
  • K (trap stiffness) = 0.455
    Used in Eq (4) to compute optical force; described as empirically scaled, not measured from a physical trap.
  • delta (near/far threshold) = 0.446
    Threshold separating linear and inverse-square force regimes in Eq (4).
  • A (far-field attraction scale) = 0.058
    Scales attractive force in far field, Eq (4).
  • C (distance offset) = 0.01
    Offset for distant interactions in Eq (4).
  • d1, d2 (shared control distance thresholds) = 1, 2
    Thresholds in Eq (5) that determine when the human or RL control dominates; chosen by hand, not derived.
  • Reward thresholds = 10 pN, 0.2 um
    Hand-chosen limits for contact force and trap-center deviation in the RL reward function (Section III-B).
assumptions (5)
  • domain assumption Isaac Sim physics engine accurately simulates micro-scale fluid dynamics, Brownian motion, and contact forces
    The platform relies on this for high-fidelity simulation, stated in Section II-B without experimental validation.
  • domain assumption Eq (4) is a valid optical trap force model
    The force feedback and simulation use Eq (4) with empirically scaled constants; no calibration data are provided (Section II-D).
  • domain assumption Haptic force scaling provides realistic and stable feedback
    The system scales forces 'empirically' and applies low-pass filters; perception is not validated against real force measurements (Section II-D).
  • domain assumption A* plus B-spline path smoothing is a suitable global planner for the RL agent
    The paper assumes this hybrid approach reduces training cost and works in dynamic environments, shown in Section II-C without comparison to alternatives.
  • ad hoc to paper The DQN agent converges to a good policy in 1000 episodes
    Training time and episode count are specific to this setup, not derived from theory (Section II-C).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Interactive OT Gym: A Reinforcement Learning-Based Interactive Optical tweezer (OT)-Driven Microrobotics Simulation Platform." pith.science (2026). https://pith.science/paper/2XP226NU

@misc{pith2026250520751,
  author       = {Pith},
  title        = {Pith review of: Interactive OT Gym: A Reinforcement Learning-Based Interactive Optical tweezer (OT)-Driven Microrobotics Simulation Platform},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2XP226NU}},
  note         = {Machine review of arXiv:2505.20751}
}
read the original abstract

Optical tweezers (OT) offer unparalleled capabilities for micromanipulation with submicron precision in biomedical applications. However, controlling conventional multi-trap OT to achieve cooperative manipulation of multiple complex-shaped microrobots in dynamic environments poses a significant challenge. To address this, we introduce Interactive OT Gym, a reinforcement learning (RL)-based simulation platform designed for OT-driven microrobotics. Our platform supports complex physical field simulations and integrates haptic feedback interfaces, RL modules, and context-aware shared control strategies tailored for OT-driven microrobot in cooperative biological object manipulation tasks. This integration allows for an adaptive blend of manual and autonomous control, enabling seamless transitions between human input and autonomous operation. We evaluated the effectiveness of our platform using a cell manipulation task. Experimental results show that our shared control system significantly improves micromanipulation performance, reducing task completion time by approximately 67% compared to using pure human or RL control alone and achieving a 100% success rate. With its high fidelity, interactivity, low cost, and high-speed simulation capabilities, Interactive OT Gym serves as a user-friendly training and testing environment for the development of advanced interactive OT-driven micromanipulation systems and control algorithms. For more details on the project, please see our website https://sites.google.com/view/otgym

Figures

Figures reproduced from arXiv: 2505.20751 by the authors.

Figure 1
Figure 1. Overview of the Interactive OT Gym Components: Haptic Devices [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The distributed architecture of shared control (connected by ROS), the motion of microrobots is determined by both RL control and human [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Progressive training process for autonomous navigation. [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Framework for human-robot shared control based on context-aware [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]
Figure 5
Figure 5. Figure 5: (a) Visualization of the three paths: manual, A*, and B-spline smoothed paths. (b) Spectral analysis results of the three paths: manual, A*, and [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: Safety evolution during RL-based robot navigation training: distri [PITH_FULL_IMAGE:figures/full_fig_p005_6.png]
Figure 7
Figure 7. Figure 7: Comparison of success rate (a) and task completion time (b) under manual, autonomous, and shared control modes. (c) NASA-TLX scores under [PITH_FULL_IMAGE:figures/full_fig_p006_7.png]
Figure 8
Figure 8. Figure 8: User interface during microrobot manipulation, showing local and [PITH_FULL_IMAGE:figures/full_fig_p006_8.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

27 extracted references · 26 canonical work pages

  1. [1]

    An integrated flow cytometry-based platform for isolation and molecular characterization of circulating tumor single cells and clusters,

    N. Bhagwat, K. Dulmage, C. H. Pletcher Jr, L. Wang, W. DeMuth, M. Sen, D. Balli, S. S. Yee, S. Sa, F. Tong et al., “An integrated flow cytometry-based platform for isolation and molecular characterization of circulating tumor single cells and clusters,” Scientific reports, vol. 8, no. 1, p. 5035, 2018

    work page 2018

  2. [2]

    A revolution in optical manipulation,

    D. G. Grier, “A revolution in optical manipulation,” nature, vol. 424, no. 6950, pp. 810–816, 2003

    work page 2003

  3. [3]

    Enhanced cell sorting and manipulation with combined optical tweezer and microfluidic chip technologies,

    X. Wang, S. Chen, M. Kong, Z. Wang, K. D. Costa, R. A. Li, and D. Sun, “Enhanced cell sorting and manipulation with combined optical tweezer and microfluidic chip technologies,” Lab on a Chip , vol. 11, no. 21, pp. 3656–3662, 2011

    work page 2011

  4. [4]

    Cell sorting with combined optical tweezers and microfluidic chip technologies,

    X. Wang, Z. Wang, and D. Sun, “Cell sorting with combined optical tweezers and microfluidic chip technologies,” in 2010 11th Interna- tional Conference on Control Automation Robotics & Vision . IEEE, 2010, pp. 201–206

    work page 2010

  5. [5]

    Versatile, facile and low-cost single-cell isolation, culture and sequencing by optical tweezer-assisted pool-screening,

    T. Xu, Y . Li, X. Han, L. Kan, J. Ren, L. Sun, Z. Diao, Y . Ji, P. Zhu, J. Xu et al., “Versatile, facile and low-cost single-cell isolation, culture and sequencing by optical tweezer-assisted pool-screening,” Lab on a Chip, vol. 23, no. 1, pp. 125–135, 2023

    work page 2023

  6. [6]

    Optical tweezers: Phototoxicity and thermal stress in cells and biomolecules,

    A. Bl ´azquez-Castro, “Optical tweezers: Phototoxicity and thermal stress in cells and biomolecules,” Micromachines, vol. 10, no. 8, p. 507, 2019

    work page 2019

  7. [7]

    Indirect optical trapping using light driven micro-rotors for reconfigurable hydrodynamic manipulation,

    U. G. B ¯utait˙e, G. M. Gibson, Y .-L. D. Ho, M. Taverne, J. M. Taylor, and D. B. Phillips, “Indirect optical trapping using light driven micro-rotors for reconfigurable hydrodynamic manipulation,” Nature communications, vol. 10, no. 1, p. 1215, 2019

    work page 2019

  8. [8]

    Manipulation of biological cells using a robot-aided optical tweezers system,

    M. Xie, A. Shakoor, and C. Wu, “Manipulation of biological cells using a robot-aided optical tweezers system,” Micromachines, vol. 9, no. 5, p. 245, 2018

    work page 2018

Show all 27 references
  1. [9]

    A review on microrobots driven by optical and magnetic fields,

    Y . Hou, H. Wang, R. Fu, X. Wang, J. Yu, S. Zhang, Q. Huang, Y . Sun, and T. Fukuda, “A review on microrobots driven by optical and magnetic fields,” Lab on a Chip , vol. 23, no. 5, pp. 848–868, 2023

    2023

  2. [10]

    Automated manipulation of biological cells using gripper formations controlled by optical tweezers,

    S. Chowdhury, A. Thakur, P. ˇSvec, C. Wang, W. Losert, and S. K. Gupta, “Automated manipulation of biological cells using gripper formations controlled by optical tweezers,” IEEE Transactions on Automation Science and Engineering , vol. 11, no. 2, pp. 338–347, 2013

    2013

  3. [11]

    Stochastic control for orientation and transportation of microscopic objects using multiple optically driven robotic fingertips,

    Q. M. Ta and C. C. Cheah, “Stochastic control for orientation and transportation of microscopic objects using multiple optically driven robotic fingertips,” IEEE Transactions on Robotics, vol. 35, no. 4, pp. 861–872, 2019

    2019

  4. [12]

    Indirect pushing based automated micromanipulation of bi- ological cells using optical tweezers,

    A. Thakur, S. Chowdhury, P. ˇSvec, C. Wang, W. Losert, and S. K. Gupta, “Indirect pushing based automated micromanipulation of bi- ological cells using optical tweezers,” The International Journal of Robotics Research, vol. 33, no. 8, pp. 1098–1111, 2014

    2014

  5. [13]

    Distributed force control for microrobot manipulation via planar multi-spot optical tweezer,

    D. Zhang, A. Barbot, B. Lo, and G.-Z. Yang, “Distributed force control for microrobot manipulation via planar multi-spot optical tweezer,” Advanced Optical Materials , vol. 8, no. 21, p. 2000543, 2020

    2020

  6. [14]

    High- bandwidth 3d force feedback optical tweezers for interactive bio- manipulation,

    M. Yin, E. Gerena, C. Pacoret, S. Haliyo, and S. Regnier, “High- bandwidth 3d force feedback optical tweezers for interactive bio- manipulation,” in 2017 IEEE/RSJ International Conference on Intel- ligent Robots and Systems (IROS) . IEEE, 2017, pp. 1889–1894

    2017

  7. [15]

    Improving optical micromanipulation with force-feedback bilateral coupling,

    E. Gerena, F. Legendre, Y . Vitry, S. R´egnier, and S. Haliyo, “Improving optical micromanipulation with force-feedback bilateral coupling,” in 2020 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2020, pp. 10 292–10 298

    2020

  8. [16]

    Dual-arm visuo-haptic optical tweezers for bimanual cooperative micromanipulation of nonspherical objects,

    Y . Tanaka and K. Fujimoto, “Dual-arm visuo-haptic optical tweezers for bimanual cooperative micromanipulation of nonspherical objects,” Micromachines, vol. 13, no. 11, p. 1830, 2022

    2022

  9. [17]

    2d high speed force feedback teleoperation of optical tweezers,

    Z. Ni, C. Pacoret, R. Benosman, and S. R ´egnier, “2d high speed force feedback teleoperation of optical tweezers,” in 2013 IEEE International Conference on Robotics and Automation . IEEE, 2013, pp. 1700–1705

    2013

  10. [18]

    Invited article: A review of haptic optical tweezers for an interactive microworld exploration,

    C. Pacoret and S. R ´egnier, “Invited article: A review of haptic optical tweezers for an interactive microworld exploration,” Review of Scientific Instruments , vol. 84, no. 8, 2013

    2013

  11. [19]

    Intuitive cell manipulation microscope system with haptic device for intracy- toplasmic sperm injection simplification,

    K. Sakamoto, T. Aoyama, M. Takeuchi, and Y . Hasegawa, “Intuitive cell manipulation microscope system with haptic device for intracy- toplasmic sperm injection simplification,” Sensors, vol. 24, no. 2, p. 711, 2024

    2024

  12. [20]

    Collaborative teleoperation with haptic feedback for collision-free navigation of ground robots,

    M. Coffey and A. Pierson, “Collaborative teleoperation with haptic feedback for collision-free navigation of ground robots,” in 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, 2022, pp. 8141–8148

    2022

  13. [21]

    Real-time teleoperation of magnetic force-driven microrobots with 3d haptic force feedback for micro-navigation and micro-transportation,

    J. Lee, X. Zhang, C. H. Park, and M. J. Kim, “Real-time teleoperation of magnetic force-driven microrobots with 3d haptic force feedback for micro-navigation and micro-transportation,” IEEE Robotics and Automation Letters, vol. 6, no. 2, pp. 1769–1776, 2021

    2021

  14. [22]

    Playing optical tweezers with deep reinforcement learning: in virtual, physical and augmented environments,

    M. Praeger, Y . Xie, J. A. Grant-Jacob, R. W. Eason, and B. Mills, “Playing optical tweezers with deep reinforcement learning: in virtual, physical and augmented environments,” Machine Learning: Science and Technology, vol. 2, no. 3, p. 035024, 2021

    2021

  15. [23]

    Machine learning-enhanced optical tweezers for defect-free rearrangement,

    Y . Lee and E. Chae, “Machine learning-enhanced optical tweezers for defect-free rearrangement,” Current Applied Physics , vol. 61, pp. 150–159, 2024

    2024

  16. [24]

    Autonomous 3d positional control of a magnetic microrobot using reinforcement learning,

    S. A. Abbasi, A. Ahmed, S. Noh, N. L. Gharamaleki, S. Kim, A. M. B. Chowdhury, J.-y. Kim, S. Pan ´e, B. J. Nelson, and H. Choi, “Autonomous 3d positional control of a magnetic microrobot using reinforcement learning,” Nature Machine Intelligence , vol. 6, no. 1, pp. 92–105, 2024

    2024

  17. [25]

    Robotic assisted micromanipulation sys- tem using virtual fixtures and metaphors,

    M. Ammi and A. Ferreira, “Robotic assisted micromanipulation sys- tem using virtual fixtures and metaphors,” in Proceedings 2007 IEEE International Conference on Robotics and Automation . IEEE, 2007, pp. 454–460

    2007

  18. [26]

    Development of a human-robot- shared controlled teletweezing system,

    G. Hwang and H. Hashimoto, “Development of a human-robot- shared controlled teletweezing system,” IEEE Transactions on Control Systems Technology, vol. 15, no. 5, pp. 960–966, 2007

    2007

  19. [27]

    Human–machine interaction control for stochastic cell manipulation systems,

    Q. M. Ta and C. C. Cheah, “Human–machine interaction control for stochastic cell manipulation systems,”Automatica, vol. 131, p. 109721, 2021

    2021

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.