Pith. sign in

REVIEW 3 major objections 5 minor 75 references

CART-MPC: Coordinating Assistive Devices for Robot-Assisted Transferring with Multi-Agent Model Predictive Control

T0 review · 3 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read CART-MPC claims a small robot arm can tie Hoyer sling straps by coordinating with an actuated sling bar, using turn-taking MPC and a linking-number cost.

desk verdict Solid, real engineering contribution with working hardware and a clear multi-agent MPC core, but the abstract oversells generalization and misuses the term 'zero-shot' for the real-world transfer. read the letter →

arxiv 2501.11149 v1 pith:FYQJKWNZ submitted 2025-01-19 cs.RO

Ruolin Ye , Shuaixing Chen , Yunting Yan , Joyce Yang , Christina Ge , Jose Barreiros , Kate Tsui , Tom Silver
show 1 more author
Tapomayukh Bhattacharjee
This is my paper · ORCID
classification cs.RO MSC 68T4093C8557K10
keywords robot-assistedtransferringmodelpredictivecontroldeformableobjectmanipulationmulti-agentcoordinationlinkingnumberkeypointdynamicssim-to-realtransferassistivedevices
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

CART-MPC claims that the bed-to-wheelchair transfer task, or at least its strap-tying subtask, does not require a heavy-lift robot: a small robot arm can fasten a deformable Hoyer sling strap to a sling bar if the sling bar itself is actuated and plans cooperatively with the arm. The paper proposes a turn-taking multi-agent model predictive controller in which each agent plans its own action while assuming the other stays still, guided by a learned neural dynamics model of the strap in keypoint space and a cost function derived from the knot-theory linking number that measures how wound the strap is around the hook. In simulation, the method generalizes across five hook shapes, three sling materials, and five care-recipient body shapes, with average single-strap success 0.82 against 0.59 and 0.54 for single-agent baselines; in real-world manikin trials it achieved 62 percent single-strap success with zero-shot sim-to-real transfer. The consequence is that off-the-shelf assistive devices, instrumented with actuators and sensors, can be treated as planning agents, opening a path toward autonomous transferring with payload-limited caregiving robots.

What carries the argument

The central object is the linking number $\mathrm{link}(\gamma_1,\gamma_2)$ of the Hoyer strap curve and the sling-bar hook curve. For closed curves it counts how many times one winds through the other; for the open hook the same intuition is used as a continuous degree-of-fastening measure. In simulation the discrete linking number is computed as a double sum over points on both curves, but it is too slow for planning and requires ground-truth pose, so the authors train a neural network to predict it from observation histories, giving $\mathrm{link}_\theta(h_t)$, and define the tying cost as $c_{\mathrm{link}}(h_t) = 1 - (\mathrm{link}_\theta(h_t)-\beta_0)/(\beta_1-\beta_0)$ normalized to $[0,1]$. This cost, together with a keypoint-space MLP dynamics model trained in simulation, is plugged into a turn-taking MPC in which robot and bar alternate one-step action selection, each assuming the other is static.

What would settle it

Place the strap directly on top of the hook without threading it through, record the controller's internal 'tied' score, and check whether the robot stops; if it stops before the strap is seated in the hook, the false-positive failure is confirmed.

Watch

Extended reading notes

Core claim

The central claim is that the problem of fastening a Hoyer sling strap to a sling bar can be decomposed into alternating decisions by two agents—the robot arm holding the strap and the actuated sling bar rotating the hook—and that a purely keypoint-based, learned model of the strap's dynamics is sufficient to plan these decisions. The load-bearing signal is the linking number between the strap curve and the hook curve, computed discretely in simulation and then amortized by a neural network so it can be evaluated from observations during planning. Minimizing the normalized cost $1 - \mathrm{link}_\theta(h_t)$ drives the strap to wind around the hook. The paper reports that this scheme outperforms single-agent MPC with an uncontrolled or fixed sling bar in simulation, generalizes to hook shapes, sling materials, and care-recipient bodies outside the training distribution, and transfers to the real world without additional training, achieving 0.62 single-strap success on a manikin.

Load-bearing premise

The algorithm's check for 'tied' is a 2D image-based estimate of how many times the strap winds around the hook; if that estimate says 'tied' when the strap is close to the hook but not actually through it, the controller stops too early.

Editorial extensions

If this is right

  • A caregiving robot with a payload too small to lift a person can still perform the strap-fastening part of a transfer by treating the sling bar as a second agent.
  • The linking-number cost gives a task-progress signal that does not depend on full 3D shape reconstruction, so deformable-object manipulation can be planned from 2D keypoints.
  • Training the dynamics model on a single sling material and hook shape is enough to generalize to other hook topologies and body shapes, though the stiffest sling material remains a reported failure case.
  • The turn-taking MPC assumption that the other agent is static during each step is sufficient in this task, avoiding the complexity of joint multi-agent optimization.

Reading between the lines

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

  • The same linking-number cost could be applied to other 'wind a strap around a hook' tasks, such as securing cargo straps, medical bandages, or wiring harnesses, as long as the topology is a curve with an opening.
  • The reported failure mode—falsely high linking number when the strap is close to but not in the hook—suggests that adding depth or 3D keypoints could improve robustness; this is a concrete next step the paper leaves implicit.
  • If the instrumented sling bar and wheelchair are accepted as platforms, the turn-taking MPC framework extends naturally to more agents, such as bed positioning and wheelchair navigation, making full bed-to-wheelchair transfer a plausible near-term goal.
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 / 5 minor

Summary. The paper presents CART-MPC, a turn-taking multi-agent model predictive control algorithm for the subtask of tying Hoyer sling straps to a sling bar during bed-to-wheelchair transfer. The robot arm and an actuated sling bar are treated as cooperating agents; a keypoint-space neural dynamics model and a knot-theory-inspired linking-number cost with neural amortization are trained in RCareWorld. In simulation, CART-MPC is evaluated on 5 hook shapes, 3 sling materials, and 5 care recipient body shapes with 50 trials per setting per method (650 trials per method), reporting average single-strap success S1 = 0.82 versus 0.59 and 0.54 for the two single-agent baselines. In the real world, the paper reports a 62% success rate over 50 single-strap tying trials on a manikin setup and demonstrates a full four-strap tying sequence. The abstract claims zero-shot sim-to-real generalization, while Section VI-C3 states that MPC parameters were tuned separately for the real-world scenario.

Significance. If the results hold, this is a meaningful contribution to robot-assisted caregiving: it shows that a lightweight robot can complete a deformable-object manipulation subtask of patient transfer by coordinating with actuated assistive devices, rather than by bearing the full payload. The simulation evaluation is extensive, with 650 trials per method, and the multi-agent advantage over single-agent baselines is supported by a z-test (p < 0.0001). The linking-number-based cost function is creative and the neural amortization is practical. However, the real-world evidence is weakened by the undisclosed per-domain controller tuning, and the abstract overstates generalization to sling materials given the sharp drop on the stiff material M3. These issues are fixable but currently affect the paper's headline claims.

major comments (3)
  1. [Abstract and Section VI-C3] The abstract's claim of 'zero-shot sim-to-real generalization capabilities' is contradicted by Section VI-C3, which states that 'we use the dynamics model obtained from simulation and transfer it directly to the real world, but tune the MPC parameters separately for the real-world scenario' and that the simulation material was tuned to align with the real world. Tuning controller parameters on the target domain is a form of sim-to-real adaptation, so the reported 62% real-world success rate does not support the zero-shot claim as stated. This is load-bearing because the abstract's strongest claim is precisely the zero-shot generalization. Please either remove the zero-shot wording or report which MPC parameters were tuned, their simulation and real-world values, and argue why this does not constitute target-domain adaptation.
  2. [Table I and Section VI-C2] The abstract's statement that CART-MPC 'successfully generalizes across diverse ... sling materials' is too strong given Table I: for the stiff sling material M3, S1 = 0.48 compared with 0.88 for the training material M1, and Section VI-C2 itself acknowledges this performance drop. Although 0.48 still outperforms the baselines (0.30 and 0.22), the wording 'successfully generalizes' should be qualified with the M3 result. In addition, Table I reports no confidence intervals; with 50 trials per cell, Wilson intervals are needed to assess whether differences such as M3 versus M1 are meaningful.
  3. [Section VII] Section VII acknowledges that 'the linking number-based cost function can sometimes produce falsely high values, resulting in tying failures, particularly when the strap is very close to the hook but not yet secured.' Because the MPC selects actions and terminates based on this cost, this is a known failure mode of the central cost function. The paper should quantify how often this failure mode occurs in the 650 simulation trials and analyze whether it explains the M3 gap or the sim-to-real drop from 0.82 to 0.62. Without this analysis, the reliability of the linking-number cost as the main task objective remains underspecified.
minor comments (5)
  1. [Section V-C4] There is a typo in Section V-C4: 'strap typing cost function' should be 'strap tying cost function.'
  2. [Section VI-A and website] The paper states that the STORM cost terms and weights are detailed on the website; for archival reproducibility, the key cost weights, prediction horizon, sampling counts, and the normalization bounds beta_0 and beta_1 should be included in the paper or a stable appendix.
  3. [Section VI-B] The full-sling tying evaluation assumes that the wheelchair's control system can accurately follow a planner-generated trajectory; the paper should clarify that the S4 and S_lift numbers measure strap tying under assumed base placement, not end-to-end navigation and manipulation autonomy.
  4. [Section V-C2] The discrete linking number formula is unnumbered and the notation M1 and M2 is introduced in a way that is easy to misread; consider adding an equation number and a sentence explicitly defining the sampled point sets on the strap and hook.
  5. [Section VI-C3] The real-world evaluation reports only a 62% success rate over 50 trials without confidence intervals or a comparison baseline; a Wilson interval and, ideally, a matched simulation condition with the same tuned parameters would substantially strengthen the sim-to-real comparison.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: CART-MPC's learned components are validated against an external physical success criterion and real-world trials, with only non-load-bearing self-citations.

full rationale

The derivation chain in CART-MPC is self-contained rather than circular. The keypoint dynamics model is trained on simulator trajectories generated by a random exploration policy, and the linking-number cost network is trained on simulator-computed linking numbers; both are then used inside a sampling-based MPC. The success criterion is defined independently of these learned components: a trial succeeds only if the strap stays securely at the hook's bottom and, in the full task, the care recipient remains stable in the sling for at least five seconds after lifting. Consequently, the MPC's minimization of clink is not definitionally equivalent to the reported success metric, and the paper even documents cases where the linking-number cost produces falsely high values and causes tying failures, confirming that the cost is a proxy rather than the ground truth being predicted. The real-world experiments, although described as 'zero-shot sim-to-real' while MPC parameters were tuned separately for the real-world scenario, are an overclaim about transfer rather than a circularity: the 62% real single-strap success rate is an external outcome, not a quantity forced by the training pipeline. The main self-citation, RCareWorld [1], is used as a simulation platform for data generation and evaluation; it is not invoked as a uniqueness theorem or as the proof of any result, so it is non-load-bearing. No equation in the paper is defined in terms of the quantity it is used to predict, and no fitted parameter is renamed as a prediction. Thus the central claim has independent empirical content, and no circular step is exhibited.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

CART-MPC rests on several modeling choices: the linking-number cost as a success proxy, keypoint state sufficiency, simulator fidelity, and the static-other-agent MPC approximation. The only numeric values fitted to data are the linking-number normalization bounds, hand-tuned MPC weights, and separately tuned real-world MPC parameters. No invented physical entities appear.

free parameters (4)
  • beta_0 and beta_1 (linking-number normalization bounds) = min and max of link_theta over training data
    Used in clink = 1 - (link_theta - beta_0)/(beta_1 - beta_0), so the cost depends on these data-derived bounds; they are computed from simulator-generated states, not from a first-principles scale.
  • MPC cost weights for STORM = not reported in paper, deferred to website
    The cost terms and weights are tuned by hand, influencing the plan; the paper states they are detailed on the website.
  • Real-world MPC parameters = retuned separately for the real world
    Sec. VI-C3 says the dynamics model transfers directly but MPC parameters are tuned for the real-world scenario, so the zero-shot claim is not parameter-free.
  • Stable-fastening duration threshold = not specified
    The goal g(o) is true when 'the strap does not fall within a fixed duration'; the duration is never quantified, so success assessment depends on an unstated threshold.
assumptions (4)
  • domain assumption The discrete linking number formula (Sec. V-C2, Eq. 1) is a valid proxy for 'strap is fastened around the hook' for open curves.
    The paper cites knot theory for closed curves and extends it to open curves; the authors report false positives near the hook, so this assumption is only approximately true.
  • domain assumption A keypoint representation of the strap in 2D pixel space contains enough information for planning the tie.
    Keypoints are obtained from segmentation and optical flow; the paper acknowledges depth is too noisy near the hook, and that the 2D-only representation causes cost false positives.
  • domain assumption RCareWorld's XPBD/Obi simulation of strap dynamics is faithful enough to transfer zero-shot to the real world.
    The dynamics model is trained entirely on simulator data; sim-to-real transfer is demonstrated on a manikin with modest success, and authors say simulator fidelity for thin objects is a remaining challenge.
  • domain assumption In turn-taking MPC, each agent may assume the other agent is static during its own prediction horizon.
    Stated in Sec. V-A: 'For simplicity, each agent assumes that the other remains static during prediction'; this decouples the MPC but ignores simultaneous motion effects.

how reviews work

0 comments
Cite this review

Pith. "Pith review of CART-MPC: Coordinating Assistive Devices for Robot-Assisted Transferring with Multi-Agent Model Predictive Control." pith.science (2026). https://pith.science/paper/FYQJKWNZ

@misc{pith2026250111149,
  author       = {Pith},
  title        = {Pith review of: CART-MPC: Coordinating Assistive Devices for Robot-Assisted Transferring with Multi-Agent Model Predictive Control},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FYQJKWNZ}},
  note         = {Machine review of arXiv:2501.11149}
}
read the original abstract

Bed-to-wheelchair transferring is a ubiquitous activity of daily living (ADL), but especially challenging for caregiving robots with limited payloads. We develop a novel algorithm that leverages the presence of other assistive devices: a Hoyer sling and a wheelchair for coarse manipulation of heavy loads, alongside a robot arm for fine-grained manipulation of deformable objects (Hoyer sling straps). We instrument the Hoyer sling and wheelchair with actuators and sensors so that they can become intelligent agents in the algorithm. We then focus on one subtask of the transferring ADL -- tying Hoyer sling straps to the sling bar -- that exemplifies the challenges of transfer: multi-agent planning, deformable object manipulation, and generalization to varying hook shapes, sling materials, and care recipient bodies. To address these challenges, we propose CART-MPC, a novel algorithm based on turn-taking multi-agent model predictive control that uses a learned neural dynamics model for a keypoint-based representation of the deformable Hoyer sling strap, and a novel cost function that leverages linking numbers from knot theory and neural amortization to accelerate inference. We validate it in both RCareWorld simulation and real-world environments. In simulation, CART-MPC successfully generalizes across diverse hook designs, sling materials, and care recipient body shapes. In the real world, we show zero-shot sim-to-real generalization capabilities to tie deformable Hoyer sling straps on a sling bar towards transferring a manikin from a hospital bed to a wheelchair. See our website for supplementary materials: https://emprise.cs.cornell.edu/cart-mpc/.

Figures

Figures reproduced from arXiv: 2501.11149 by the authors.

Figure 1
Figure 1. Bed-to-wheelchair Transferring: Human caregivers use assistive devices to perform transferring. For example, the caregiver in the figure (upper row) uses a Hoyer sling to move the care recipient from the hospital bed to the wheelchair. Inspired by this, we use an instrumented Hoyer sling along with a caregiving robot to perform the transferring task (lower row), where the robot performs fine manipulation, and the as… view at source ↗
Figure 2
Figure 2. Instrumented assistive devices: We added sensors and actuators to a Hoyer sling and a commercial powered wheelchair with a robot arm. III. INSTRUMENTED ASSISTIVE DEVICES FOR TRANSFER In this section, we briefly describe our novel hardware setup to provide context for the technical contributions that are the focus of this work. Further details can be found on our website. As illustrated in [PITH_FULL_IMAGE:figures/f… view at source ↗
Figure 3
Figure 3. (a) CART-MPC: We propose a turn-taking multi-agent algorithm to coordinate the robot and the Hoyer sling bar during strap fastening. The algorithm leverages a dynamics model for the strap and a neural cost function. See text for details. (b) Neural linking number: We visualize the average neural linking number across five strap-tying trials. The neural linking number gradually increases from 0 to 1 as the robot ties… view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Setup and results for evaluation of CART-MPC in RCareWorld: We evaluate our method using the setup on the left with various Hook shapes (H), Care Recipients (CR) with various body shapes, and Sling Materials (M) with three levels of compliance. We show the results S1 i…
Figure 5
Figure 5. Figure 5: CART-MPC executed sequence in RCareWorld: We demonstrate one executed sequence with H1, M1, and CR3. In this trial, the robot and the sling bar collaboratively fasten the 4 straps to the sling bar hooks. TABLE I: Evaluating CART-MPC: We compare our method with the base…
Figure 6
Figure 6. Figure 6: CART-MPC execute sequence in the real world: We demonstrate one execute sequence with a manikin. The robot and the Hoyer sling bar collaboratively fasten four straps. See our website for the full video. However, active control of the sling bar proves more efficient and…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

75 extracted references · 69 canonical work pages

  1. [1]

    RCareWorld: A human-centric sim- ulation world for caregiving robots,

    R. Ye, W. Xu, H. Fu, R. K. Jenamani, V . Nguyen, C. Lu, K. Dim- itropoulou, and T. Bhattacharjee, “RCareWorld: A human-centric sim- ulation world for caregiving robots,” in 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , pp. 33–40, IEEE, 2022

    work page 2022

  2. [2]

    W. H. Organization, Global report on health equity for persons with disabilities. World Health Organization, 2022

    work page 2022

  3. [3]

    Flair: Feeding via long-horizon acquisition of realistic dishes,

    R. K. Jenamani, P. Sundaresan, M. Sakr, T. Bhattacharjee, and D. Sadigh, “Flair: Feeding via long-horizon acquisition of realistic dishes,” arXiv preprint arXiv:2407.07561, 2024

    arXiv 2024

  4. [4]

    An adaptable, safe, and portable robot-assisted feeding system,

    E. K. Gordon, R. K. Jenamani, A. Nanavati, Z. Liu, D. Stabile, X. Dai, T. Bhattacharjee, T. Schrenk, J. Ko, H. Bolotski, et al., “An adaptable, safe, and portable robot-assisted feeding system,” in Companion of the 2024 ACM/IEEE International Conference on Human-Robot Interaction, pp. 74–76, 2024

    work page 2024

  5. [5]

    Human-robot commensality: Bite timing prediction for robot-assisted feeding in groups,

    J. Ondras, A. Anwar, T. Wu, F. Bu, M. Jung, J. J. Ortiz, and T. Bhattacharjee, “Human-robot commensality: Bite timing prediction for robot-assisted feeding in groups,” in 6th Annual Conference on Robot Learning, 2022

    2022

  6. [6]

    Repeat: A real2sim2real approach for pre-acquisition of soft food items in robot-assisted feeding,

    H. Nayoung, Y . Ruolin, L. Ziang, S. Shubhangi, and B. Tapomayukh, “Repeat: A real2sim2real approach for pre-acquisition of soft food items in robot-assisted feeding,” 2024

    work page 2024

  7. [7]

    Feel the bite: Robot-assisted inside-mouth bite transfer using robust mouth perception and physical interaction-aware control,

    R. K. Jenamani, D. Stabile, Z. Liu, A. Anwar, K. Dimitropoulou, and T. Bhattacharjee, “Feel the bite: Robot-assisted inside-mouth bite transfer using robust mouth perception and physical interaction-aware control,” in ACM/IEEE International Conference on Human Robot Interaction (HRI), 2024

    work page 2024

  8. [8]

    To ask or not to ask: Human-in-the-loop contextual bandits with applications in robot-assisted feeding,

    R. Banerjee, R. K. Jenamani, S. Vasudev, A. Nanavati, S. Dean, and T. Bhattacharjee, “To ask or not to ask: Human-in-the-loop contextual bandits with applications in robot-assisted feeding,” 2024

    work page 2024

Show all 75 references
  1. [9]

    Morpheus: a multimodal one-armed robot-assisted peeling system with human users in-the-loop,

    Y . Ruolin, H. Yifei, B. Yuhan, K. Luke, and B. Tapomayukh, “Morpheus: a multimodal one-armed robot-assisted peeling system with human users in-the-loop,” 2024

    2024

  2. [10]

    Rabbit: A robot-assisted bed bathing system with multimodal perception and integrated compliance,

    R. Madan, S. Valdez, D. Kim, S. Fang, L. Zhong, D. T. Virtue, and T. Bhattacharjee, “Rabbit: A robot-assisted bed bathing system with multimodal perception and integrated compliance,” in Proceedings of the 2024 ACM/IEEE International Conference on Human-Robot Interaction, pp. ...

    2024

  3. [11]

    Towards an assistive robot that autonomously performs bed baths for patient hygiene,

    C.-H. King, T. L. Chen, A. Jain, and C. C. Kemp, “Towards an assistive robot that autonomously performs bed baths for patient hygiene,” in2010 IEEE/RSJ International Conference on Intelligent Robots and Systems , pp. 319–324, IEEE, 2010

    2010

  4. [12]

    Personalized robot assistant for support in dressing,

    A. Jevti ´c, A. F. Valle, G. Aleny `a, G. Chance, P. Caleb-Solly, S. Do- gramadzi, and C. Torras, “Personalized robot assistant for support in dressing,” IEEE transactions on cognitive and developmental systems , vol. 11, no. 3, pp. 363–374, 2018

    2018

  5. [13]

    Do you need a hand?– a bimanual robotic dressing assistance scheme,

    J. Zhu, M. Gienger, G. Franzese, and J. Kober, “Do you need a hand?– a bimanual robotic dressing assistance scheme,” IEEE Transactions on Robotics, vol. 40, pp. 1906–1919, 2024

    1906

  6. [14]

    RIKEN-TRI Collaboration Center for Human-Interactive Robot Re- search, “Ri-man,” 2006

    2006

  7. [15]

    Stability analysis of electrical powered wheelchair-mounted robotic-assisted transfer device,

    H. Wang, C.-Y . Tsai, H. Jeannis, C. S. Chung, A. Kelleher, G. G. Grindle, and R. A. Cooper, “Stability analysis of electrical powered wheelchair-mounted robotic-assisted transfer device,” Journal of Reha- bilitation Research and Development, vol. 51, no. 5, pp. 761–774, 2014

    2014

  8. [16]

    Evaluation of the efficacy of a lift-assist device regarding caregiver posture and muscle load for transferring tasks,

    Y .-K. Kong, K.-H. Choi, S.-S. Park, J.-W. Shim, and H.-H. Shim, “Evaluation of the efficacy of a lift-assist device regarding caregiver posture and muscle load for transferring tasks,” International Journal of Environmental Research and Public Health , vol. 20, no. 2, p. 1174, 2023

    2023

  9. [17]

    Model-based control with sparse neural dynamics,

    Z. Liu, G. Zhou, J. He, T. Marcucci, L. Fei-Fei, J. Wu, and Y . Li, “Model-based control with sparse neural dynamics,” in Thirty-seventh Conference on Neural Information Processing Systems (NeurIPS), (Van- couver, Canada), December 2023

    2023

  10. [18]

    Sam 2: Segment anything in images and videos,

    N. Ravi, V . Gabeur, Y .-T. Hu, R. Hu, C. Ryali, T. Ma, H. Khedr, R. R ¨adle, C. Rolland, L. Gustafson, E. Mintun, J. Pan, K. V . Alwala, N. Carion, C.-Y . Wu, R. Girshick, P. Doll´ar, and C. Feichtenhofer, “Sam 2: Segment anything in images and videos,” 2024

    2024

  11. [19]

    Seg- ment anything meets point tracking,

    F. Raji ˇc, L. Ke, Y .-W. Tai, C.-K. Tang, M. Danelljan, and F. Yu, “Seg- ment anything meets point tracking,” arXiv preprint arXiv:2307.01197 , 2023

    2023 arXiv

  12. [20]

    Knot polynomials of open and closed curves,

    E. Panagiotou and L. H. Kauffman, “Knot polynomials of open and closed curves,” Proceedings of the Royal Society A , vol. 476, no. 2240, p. 20200124, 2020

    2020

  13. [21]

    Realization of ageing-friendly smart home system with computational intelligence,

    C.-Q. Zhang, Y .-F. Liu, and D.-P. Liang, “Realization of ageing-friendly smart home system with computational intelligence,” Journal of Com- putational Information Systems , vol. 10, no. 8, pp. 3583–3590, 2014

    2014

  14. [22]

    RIKEN-TRI Collaboration Center for Human-Interactive Robot Re- search, “Robear,” 2015

    2015

  15. [23]

    Riba: Robot for interactive body assistance,

    RIKEN-TRI Collaboration Center for Human-Interactive Robot Re- search, “Riba: Robot for interactive body assistance,” 2008

    2008

  16. [24]

    Riken-tri collaboration center for human-interactive robot research,

    RIKEN-TRI, “Riken-tri collaboration center for human-interactive robot research,” 2011. Accessed: 2024-09-26

    2011

  17. [25]

    Hlpr chair: a novel patient transfer device,

    R. Bostelman, J. Albus, and J. Johnson, “Hlpr chair: a novel patient transfer device,” in Proceedings of the 8th Workshop on Performance Metrics for Intelligent Systems , PerMIS ’08, (New York, NY , USA), p. 302–305, Association for Computing Machinery, 2008

    2008

  18. [26]

    Docking control of holonomic omnidi- rectional vehicles with applications to a hybrid wheelchair/bed system,

    S. Mascaro and H. H. Asada, “Docking control of holonomic omnidi- rectional vehicles with applications to a hybrid wheelchair/bed system,” in Proceedings. 1998 IEEE International Conference on Robotics and Automation (Cat. No. 98CH36146) , vol. 1, pp. 399–405, IEEE, 1998

    1998

  19. [27]

    Design of a new multifunctional wheelchair-bed,

    F. Li, C. Zhang, H. Liu, L. Gao, J. Ye, and D. Xin, “Design of a new multifunctional wheelchair-bed,” in World Congress on Medical Physics and Biomedical Engineering May 26-31, 2012, Beijing, China , pp. 1342–1345, Springer, 2013

    2012

  20. [28]

    Design and development of iot enabled hybrid wheelchair cum bed,

    S. Singh, S. N. Panda, R. K. Kaushal, N. Kumar, and J. L. Raheja, “Design and development of iot enabled hybrid wheelchair cum bed,” in 2021 International Conference on Emerging Smart Computing and Informatics (ESCI), pp. 711–715, IEEE, 2021

    2021

  21. [29]

    Exoskeletons in nursing and healthcare: A bionic future,

    S. O’Connor, “Exoskeletons in nursing and healthcare: A bionic future,” Clinical Nursing Research , vol. 30, pp. 1123–1126, Nov 2021. Epub 2021 Aug 8

    2021

  22. [30]

    Assistive exoskeleton for paralyzed people,

    J. Khan, K. M. R. Songlap, A. Mizan, M. Sahar, and S. Ahmed, “Assistive exoskeleton for paralyzed people,” pp. 474–479, 01 2019

    2019

  23. [31]

    Robotic exoskeletons: The current pros and cons,

    A. S. Gorgey, “Robotic exoskeletons: The current pros and cons,” World Journal of Orthopedics , vol. 9, pp. 112–119, Sep 18 2018

    2018

  24. [32]

    Ceiling lifts,

    S. Cross, “Ceiling lifts,” 2023. Accessed: September 26, 2024

    2023

  25. [33]

    Ceiling lift vs. floor lift: Pros and cons,

    P. Mobility, “Ceiling lift vs. floor lift: Pros and cons,” 2023. Accessed: September 26, 2024

    2023

  26. [34]

    Patient lift,

    W. contributors, “Patient lift,” 2023. Accessed: September 26, 2024

    2023

  27. [35]

    Guide to choosing the best hoyer lift,

    Caring.com, “Guide to choosing the best hoyer lift,” 2023. Accessed: September 26, 2024

    2023

  28. [36]

    Dudgeon, J

    B. Dudgeon, J. Deitz, and M. Dimpfel, Wheelchair Selection, pp. 495–

  29. [37]

    Smart wheelchairs: A literature review,

    R. Simpson, “Smart wheelchairs: A literature review,” Journal of reha- bilitation research and development , vol. 42, pp. 423–36, 07 2005

    2005

  30. [38]

    Achieving global co- herence in multi-agent caregiver systems: Centralized versus distributed response coordination in i.l.s.a.,

    T. A. Wagner, J. Phelps, M. Scheutz, and T. Bauer, “Achieving global co- herence in multi-agent caregiver systems: Centralized versus distributed response coordination in i.l.s.a.,” in Proceedings of the AAAI Conference on Artificial Intelligence , 2002

    2002

  31. [39]

    A multi-agent framework for coordination of intelligent assistive technologies,

    P. Valente, S. Hossain, B. Grønbæk, K. Hallenborg, and L. P. Reis, “A multi-agent framework for coordination of intelligent assistive technologies,” in 5th Iberian Conference on Information Systems and Technologies, pp. 1–6, 2010

    2010

  32. [40]

    A multirobot system in an assisted home environment to support the elderly in their daily lives,

    R. Barber, F. J. Ortiz, S. Garrido, F. M. Calatrava-Nicol ´as, A. Mora, A. Prados, J. A. Vera-Repullo, J. Roca-Gonz ´alez, I. M ´endez, and O. M. Mozos, “A multirobot system in an assisted home environment to support the elderly in their daily lives,” Sensors, vol. 22, no. 20,...

    2022

  33. [41]

    A multi-agent system based on reactive decision rules for solving the caregiver routing problem in home health care,

    E. Marcon, S. Chaabane, Y . Sallez, T. Bonte, and D. Trentesaux, “A multi-agent system based on reactive decision rules for solving the caregiver routing problem in home health care,” Simulation Modelling Practice and Theory , vol. 74, pp. 134–151, May 2017. IF=1.954

    2017

  34. [42]

    Robocare: an integrated robotic system for the domestic care of the elderly,

    S. Bahadori, A. Cesta, G. Grisetti, L. Iocchi, G. R. Leone, D. Nardi, A. Oddi, F. Pecora, and R. Rasconi, “Robocare: an integrated robotic system for the domestic care of the elderly,” 01 2003

    2003

  35. [43]

    Klonovs, M

    J. Klonovs, M. A. Haque, V . Krueger, K. Nasrollahi, K. Andersen- Ranberg, T. B. Moeslund, and E. G. Spaich, Distributed computing and monitoring technologies for older patients . Springer, 2015

    2015

  36. [44]

    Decentralized collision avoidance, deadlock detection, and deadlock resolution for multiple mobile robots,

    M. Jager and B. Nebel, “Decentralized collision avoidance, deadlock detection, and deadlock resolution for multiple mobile robots,” in Proceedings 2001 IEEE/RSJ International Conference on Intelligent Robots and Systems. Expanding the Societal Role of Robotics in the the Next ...

    2001

  37. [45]

    Planning for decentralized control of multiple robots under uncertainty,

    C. Amato, G. Konidaris, G. Cruz, C. A. Maynor, J. P. How, and L. P. Kaelbling, “Planning for decentralized control of multiple robots under uncertainty,” in 2015 IEEE international conference on robotics and automation (ICRA), pp. 1241–1248, IEEE, 2015

    2015

  38. [46]

    Decentralized task and path planning for multi-robot systems,

    Y . Chen, U. Rosolia, and A. D. Ames, “Decentralized task and path planning for multi-robot systems,” IEEE Robotics and Automation Letters, vol. 6, no. 3, pp. 4337–4344, 2021

    2021

  39. [47]

    Multi-agent model predictive control: A survey,

    R. R. Negenborn, B. De Schutter, and J. Hellendoorn, “Multi-agent model predictive control: A survey,” arXiv preprint arXiv:0908.1076 , 2009

    2009 arXiv

  40. [48]

    Trajectory opti- mization for nonlinear multi-agent systems using decentralized learning model predictive control,

    E. L. Zhu, Y . R. St ¨urz, U. Rosolia, and F. Borrelli, “Trajectory opti- mization for nonlinear multi-agent systems using decentralized learning model predictive control,” in 2020 59th IEEE Conference on Decision and Control (CDC) , pp. 6198–6203, IEEE, 2020

    2020

  41. [49]

    Decentralized multi-agent planning using model predictive control and time-aware safe corridors,

    C. Toumieh and A. Lambert, “Decentralized multi-agent planning using model predictive control and time-aware safe corridors,” IEEE Robotics and Automation Letters , vol. 7, no. 4, pp. 11110–11117, 2022

    2022

  42. [50]

    Clothpose: A real-world benchmark for visual analysis of garment pose via an indi- rect recording solution,

    W. Xu, W. Du, H. Xue, Y . Li, R. Ye, Y .-F. Wang, and C. Lu, “Clothpose: A real-world benchmark for visual analysis of garment pose via an indi- rect recording solution,” in Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) , pp. 58–68, October 2023

    2023

  43. [51]

    Gar- menttracking: Category-level garment pose tracking,

    H. Xue, W. Xu, J. Zhang, T. Tang, Y . Li, W. Du, R. Ye, and C. Lu, “Gar- menttracking: Category-level garment pose tracking,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 21233–21242, June 2023

    2023

  44. [52]

    Unifolding: Towards sample-efficient, scalable, and generalizable robotic garment folding,

    H. Xue, Y . Li, W. Xu, H. Li, D. Zheng, and C. Lu, “Unifolding: Towards sample-efficient, scalable, and generalizable robotic garment folding,” in CoRL Conference on Robot Learning , 2023

    2023

  45. [53]

    One policy to dress them all: Learning to dress people with diverse poses and garments,

    Y . Wang, Z. Sun, Z. Erickson, and D. Held, “One policy to dress them all: Learning to dress people with diverse poses and garments,” in Robotics: Science and Systems (RSS) , 2023

    2023

  46. [54]

    Repeat: A real2sim2real approach for pre-acquisition of soft food items in robot- assisted feeding,

    N. Ha, R. Ye, Z. Liu, S. Sinha, and T. Bhattacharjee, “Repeat: A real2sim2real approach for pre-acquisition of soft food items in robot- assisted feeding,” in IEEE International Conference on Intelligent Robots and Systems , 2024

    2024

  47. [55]

    MORPHeus: a Multimodal One-armed Robot-assisted Peeling system with Human Users in-the-loop,

    R. Ye, Y . Hu, Y . Bian, L. Kulm, and T. Bhattacharjee, “MORPHeus: a Multimodal One-armed Robot-assisted Peeling system with Human Users in-the-loop,” in International Conference on Robotics and Au- tomation, 2024

    2024

  48. [56]

    Bottom dressing by a life-sized humanoid robot provided failure detection and recovery functions,

    K. Yamazaki, R. Oya, K. Nagahama, K. Okada, and M. Inaba, “Bottom dressing by a life-sized humanoid robot provided failure detection and recovery functions,” in Proceedings of the 2014 IEEE/SICE Interna- tional Symposium on System Integration, (Tokyo, Japan), December 13– 15 2014

    2014

  49. [57]

    Design of service robot for wearing and taking off footwear,

    Y . Li, A. Xiao, Q. Feng, T. Zou, and C. Tian, “Design of service robot for wearing and taking off footwear,” in E3S Web of Conferences, vol. 189, p. 03024, 2020

    2020

  50. [58]

    Speedfolding: Learning efficient bimanual folding of garments,

    Y . Avigal, L. Berscheid, T. Asfour, T. Kroger, and K. Goldberg, “Speedfolding: Learning efficient bimanual folding of garments,” 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp. 1–8, 2022

    2022

  51. [59]

    A method for a compliant robot arm to perform a bandaging task on a swaying arm: A proposed approach,

    J. Li, W. Sun, X. Gu, J. Guo, J. Ota, Z. Huang, and Y . Zhang, “A method for a compliant robot arm to perform a bandaging task on a swaying arm: A proposed approach,” IEEE Robotics and Automation Magazine , vol. 30, pp. 50–61, 2023

    2023

  52. [60]

    Learning predictive representations for deformable objects using contrastive estimation,

    W. Yan, A. Vangipuram, P. Abbeel, and L. Pinto, “Learning predictive representations for deformable objects using contrastive estimation,” in Conference on Robot Learning , pp. 564–574, PMLR, 2021

    2021

  53. [61]

    Dsqnet: a deformable model-based supervised learning algorithm for grasping unknown occluded objects,

    S. Kim, T. Ahn, Y . Lee, J. Kim, M. Y . Wang, and F. C. Park, “Dsqnet: a deformable model-based supervised learning algorithm for grasping unknown occluded objects,” IEEE Transactions on Automation Science and Engineering, vol. 20, no. 3, pp. 1721–1734, 2022

    2022

  54. [62]

    Unsupervised learning for physical interaction through video prediction,

    C. Finn, I. Goodfellow, and S. Levine, “Unsupervised learning for physical interaction through video prediction,” in Advances in Neural Information Processing Systems , vol. 29, 2016

    2016

  55. [63]

    Adam: A method for stochastic optimization,

    D. P. Kingma, “Adam: A method for stochastic optimization,” arXiv preprint arXiv:1412.6980, 2014

    2014 arXiv

  56. [64]

    Rovi x3 power wheelchair,

    Rovi Mobility, “Rovi x3 power wheelchair,” 2024. Accessed: 2024-01- 05

    2024

  57. [65]

    Kinova Robotics, 2019

    Kinova Inc., Gen3 Ultra lightweight robot - Technical Documentation . Kinova Robotics, 2019. Accessed: 2024-01-05

    2019

  58. [66]

    2f-85 adaptive robot gripper,

    Robotiq Inc., “2f-85 adaptive robot gripper,” 2023

    2023

  59. [67]

    Invacare reliant 450 electric patient lift,

    Invacare Corporation, “Invacare reliant 450 electric patient lift,” 2024. Accessed: 2024-01-05

    2024

  60. [68]

    Dynamixel mx-64t,

    ROBOTIS, “Dynamixel mx-64t,” 2023

    2023

  61. [69]

    Intel realsense lidar camera l515,

    Intel Corporation, “Intel realsense lidar camera l515,” 2023

    2023

  62. [70]

    Intel realsense depth camera d415,

    Intel Corporation, “Intel realsense depth camera d415,” 2023

    2023

  63. [71]

    Intel realsense depth camera d435,

    Intel Corporation, “Intel realsense depth camera d435,” 2023

    2023

  64. [72]

    Arduino due,

    Arduino, “Arduino due,” 2023

    2023

  65. [73]

    Raspberry pi,

    Raspberry Pi Foundation, “Raspberry pi,” 2023

    2023

  66. [74]

    STORM: An integrated framework for fast joint-space model-predictive control for reactive manipulation,

    M. Bhardwaj, B. Sundaralingam, A. Mousavian, N. D. Ratliff, D. Fox, F. Ramos, and B. Boots, “STORM: An integrated framework for fast joint-space model-predictive control for reactive manipulation,” in Conference on Robot Learning , pp. 750–759, PMLR, 2022

    2022

  67. [75]

    RTAB-Map as an open-source lidar and visual simultaneous localization and mapping library for large-scale and long-term online operation,

    M. Labb ´e and F. Michaud, “RTAB-Map as an open-source lidar and visual simultaneous localization and mapping library for large-scale and long-term online operation,” Journal of Field Robotics , vol. 36, no. 2, pp. 416–446, 2019

    2019

Pith tools

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