Pith. sign in

REVIEW 2 major objections 3 minor 31 references

A control scheme for collaborative object transportation between a human and a quadruped robot using the MIGHTY suction cup

T0 review · 2 major / 3 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read Quadruped and human carry objects with a suction cup that never detaches.

desk verdict Real integration work, but the passivity proof for no-detachment has a genuine gap and the experiments are single-shot. read the letter →

arxiv 2508.00584 v1 pith:S6AI2VKT submitted 2025-08-01 cs.RO

Konstantinos Plotas , Emmanouil Papadakis , Drosakis Drosakis , Panos Trahanias , Dimitrios Papageorgiou This is my paper
classification cs.RO
keywords human-robotcollaborationadmittancecontrolquadrupedrobotsuctioncupbarrierartificialpotentialpassivityobjecttransportationforce/torquesensing
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper proposes an admittance-based control scheme for a human and a quadruped robot transporting an object together, using the MIGHTY suction cup both as gripper and force/torque sensor. The controller varies its damping according to the power the human transfers to the robot, aiming to reduce physical effort and improve controllability. To prevent the object from detaching, it adds a control signal derived from a barrier artificial potential that acts as a virtual torque around the robot's vertical axis. The authors prove the closed-loop system is passive and that the holding force criterion holds for all time if the human input has bounded energy, and they demonstrate both functions on a Unitree Go1.

What carries the argument

The load-bearing device is the barrier artificial potential $W(f_m)$ with $f_m = h(f_s) - f_{\min}$, where $h$ is the exponential smooth minimum (12) of the four chamber forces. The control applies the gradient of $W$ with respect to the robot's rotation angle $\theta$ as an extra torque around the $z$-axis, $F_v = [0,0,-\partial W/\partial \theta]^\top$. Because $h$ is always at most the true minimum, keeping $f_m > 0$ guarantees the holding criterion $\min(f_s) > f_{\min}$. The passivity argument uses the storage function $L = \tfrac{1}{2} x_2^\top M_d x_2 + \tfrac{1}{2} m_\theta x_3^2 + W$ and the completing-the-squares bound (22) to show $W$ stays bounded when the input energy is bounded.

What would settle it

Set up a carrying task in which the human pulls the object straight away from the suction cup (normal to the cup face) without rotating the robot, while the controller is operating with the barrier potential enabled. If the object detaches in continuous time or if the measured storage function $L$ grows without bound despite bounded input energy, the theorem's guarantee is falsified. Equivalently, measure $dW/dt$ directly during such a pull and compare it with $(\partial W/\partial \theta)\dot{\theta}$; a large discrepancy shows the proof's key approximation is violated.

Watch

Extended reading notes

Core claim

The central claim is that the closed-loop admittance system (17), which includes the variable damping law (6) and the barrier-potential torque (9), is passive with respect to the velocity and the human-applied input, and that as a consequence the minimum estimated chamber force never falls below the detachment threshold $f_{\min}$. The proof uses the storage function $L$ equal to the kinetic energy of the target admittance model plus the barrier potential $W$; passivity bounds $L$ by the input energy, and since $W$ grows without bound as the smooth-minimum margin $f_m = h(f_s) - f_{\min}$ approaches zero, the bound on $L$ keeps $f_m$ positive and the object attached.

Load-bearing premise

The proof that the no-detachment guarantee holds assumes that the barrier potential $W$ changes almost entirely through the robot's rotation angle $\theta$, so that the time derivative of $W$ can be replaced by $(\partial W/\partial \theta)\dot{\theta}$; in reality the human's applied force directly changes the four chamber forces and therefore $W$, and if that direct effect dominates, the passivity inequality and the boundedness of $W$ no longer follow.

Editorial extensions

If this is right

  • In continuous time, with bounded-energy human input, the object will never detach from the suction cup because the smooth-minimum margin $f_m$ stays positive (Theorem 1).
  • The variable damping factor increases when the human pushes against the motion, making the system easy to stop, and decreases when the human drives forward, lowering the energy cost.
  • Because $h(f_s) \leq \min(f_s)$, the controller acts earlier than the actual detachment boundary, giving a safety margin that can be tuned with gains $k_1$ and $k_2$.
  • The suction cup's four chamber pressures are enough to estimate the interaction force/torque, so the scheme needs no additional sensor hardware.

Reading between the lines

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

  • In the paper's proof, the approximation $dW/dt \approx (\partial W/\partial \theta)\, \dot{\theta}$ is the argument's linchpin; a rigorous check of when human forcing makes the other terms non-negligible would determine how broadly the no-detachment guarantee applies.
  • Since the barrier control signal only produces a torque about the $z$-axis, a threat of detachment caused by a straight pull away from the cup might not be fully counteracted; a translational barrier component would be a natural extension.
  • The 10 Hz force-sensor rate leaves a gap between the continuous-time guarantee and the discrete-time implementation; the authors note this, and it suggests a testable prediction that higher sampling rates shrink the detachment risk.
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

2 major / 3 minor

Summary. The paper proposes an admittance-control scheme for a quadruped robot (Unitree Go1) carrying an object with the MIGHTY suction cup, which simultaneously grasps the object and serves as a force/torque sensor. The controller combines a power-based variable damping term, adapted from prior work, with a barrier artificial potential that adds a torque around the vertical axis in order to prevent detachment of the object. The authors state a passivity theorem and a continuous-time no-detachment guarantee, and they report two experiments: one in which the robot walks along an arc with and without the barrier potential, and one in which the object is translated with constant low, constant high, and variable damping.

Significance. If the theorem and experiments were fully substantiated, the work would be a useful step toward practical physical human-robot collaboration with legged robots: it integrates a sensing-and-grasping suction cup with a variable-admittance controller, and it openly provides the code and hardware details. The algebraic passivity argument in Theorem 1 is clean and easy to follow, and the comparison of BAP-enabled versus BAP-disabled behavior, plus the two-level damping comparison, targets the right questions. However, the central safety guarantee currently rests on an unproven derivative approximation, and the experimental evaluation is a single-trial demonstration without statistical support. The significance is therefore conditional on repairing the proof (or weakening the claim) and on strengthening the experimental evidence.

major comments (2)
  1. [Theorem 1 / Eqs. (18)-(22)] The proof of Argument 2 assumes dW/dt ≈ (∂W/∂x1) ẋ1 immediately after Eq. (18), and this assumption is load-bearing. The cancellation that turns Eq. (20) into the passivity inequality and the subsequent boundedness of L in Eq. (22) require the exact identity ∂W/∂θ = (∂W/∂fs)(∂fs/∂θ) with ∂fs/∂θ given by Eq. (16). But fs is the measured four-chamber force vector, which is directly affected by the human-applied wrench and by load redistribution; it is not a function of the state x in Eq. (17). Its time derivative contains terms that are independent of θdot (for example, a sustained human push can change the chamber forces at constant θ). In that case Eq. (20) has uncompensated terms, L can grow, and the inference min(fs) > fmin for all t is not established. Because W depends on the input u, L is not a storage function of the state alone, so the passivity statement is not well-posed as written. Remark 4 restricts the claim in discrete time, but the continuous-time guarantee in Theorem 1 remains unsupported unless the approximation is replaced by a justified exact model or a new assumption is introduced.
  2. [Section V / Figs. 7-9] The experimental evaluation is based on a single trial per condition with one human operator and no error bars or statistical comparisons. The claimed reduction of both cognitive and physical load (contribution 3 and the conclusion) is therefore not substantiated by the data. In particular, the '2.6 times higher' energy result in Fig. 9b is a single-run ratio, and cognitive load is inferred qualitatively from the presence or absence of oscillatory motion rather than measured. The detachment experiment also compares a single BAP-enabled trajectory with a single BAP-disabled trajectory; given that the paper makes a safety-critical guarantee, repeated trials with varied human behavior are needed.
minor comments (3)
  1. [Eq. (7), Sec. IV] The symbols ζ and ζ are visually nearly identical; please rename the minimum and maximum damping gains (e.g., ζ_min and ζ_max) to make Eq. (7) and the experimental parameter values unambiguous.
  2. [Sec. III, Eq. (1)] The vacuum pressure is stated as '400 mPa' (0.4 Pa), which is implausible for suction holding; this is likely a typo for 400 mbar and should be corrected.
  3. [Fig. 7 and Fig. 9] The axes of Figures 7 and 9 are not labeled and the legends do not identify which curve corresponds to which chamber or which damping condition; please add units and legends.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: self-citations are method sources, not forced premises; the barrier-potential proof is a standard Lyapunov argument, with only an unvalidated derivative approximation as a correctness risk.

full rationale

The derivation chain is not circular. The variable-damping admittance law is adopted from prior work [3], the MIGHTY suction-cup sensing from [29], and the detachment threshold fmin from [30]; these are method/hardware sources with independent experimental grounding, and none is used as an axiom that forces the paper's conclusions. The barrier function W in (10) is defined in terms of the measured chamber forces and fmin, and the no-detachment guarantee in Theorem 1 is obtained by proving that W is bounded via the passivity inequality (20)-(22); this is a standard barrier/Lyapunov implication, not an equality between conclusion and definition. The proof does rely on an explicit approximation, dW/dt ≈ ∂W/∂x1 ẋ1 after (18), and if that approximation fails when human-induced chamber-force changes do not track θ_dot, inequality (20) and the boundedness of W may fail; that is a proof-validity risk, not a circularity, because W is not defined so that the conclusion holds by construction and the calibration constants plus fmin are measured independently. Remark 4 also concedes discretization limits, which narrows the continuous-time guarantee but does not make it circular. The heavy self-citation to [3], [29], [30], [31] is contextual reliance on prior same-group work, not an imported uniqueness theorem or a fitted-input-called-prediction step.

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

The central result (non-detachment guarantee and passivity) rests on the sensor calibration, the barrier potential tuning parameters, the fmin measurement, and the assumption that W is dominated by theta. No new physical entities are introduced.

free parameters (6)
  • sensor calibration parameters a_i, b_i, c = not reported numerically
    Calibrated using ground truth measurements in Eq. (1); these convert raw chamber pressures to chamber forces f_i.
  • barrier potential gains k1, k2 = k1 = 10, k2 = 1
    Tunable gains in Eq. (10); authors say k1 can be arbitrary and k2 adjusts smoothness of the control signal.
  • barrier activation threshold f0 = 20 N
    Defines the region of effect of the barrier potential in Eq. (10).
  • detachment force bound fmin = -48 N
    Measured lower bound at which the object detaches from the suction cup; used in criterion (8).
  • force-angle model parameters beta and kappa = beta = 0.04 m, kappa = 20.0 N/m
    Distance of chambers from center and stiffness coefficient in Eq. (16); not calibrated, chosen by hand.
  • admittance and variable damping parameters Md, mtheta, D1, dtheta, zeta_min, zeta_max, lambda = Md = 13I2 kg, mtheta = 1.5 kgm^2, Dd = diag(20,20,5), zeta_min=0.1, zeta_max=1, lambda=3; in translation experiment…
    Predefined target impedance and variable damping parameters chosen by the designer in Section V; these affect the human effort and controllability results.
assumptions (5)
  • domain assumption The quadruped's low-level controller accepts task-space velocity commands and compensates gravitational loads on the z-axis.
    Remark 1 states that gravitational forces from the object are compensated by the robot's low-level stabilization.
  • domain assumption The force estimation model (1)-(2) is accurate after calibration.
    The control scheme relies on measured chamber forces f_i; if calibration is biased, the barrier potential and passivity analysis are affected.
  • domain assumption The barrier potential W varies predominantly with the robot's rotation angle theta, so dW/dt is approximately (dW/dtheta) * dtheta/dt.
    Stated in the proof of Theorem 1; this approximation is necessary for the passivity inequality and the boundedness of W.
  • ad hoc to paper The relationship between chamber force changes and rotation angle is given by Eq. (16) with the sign pattern [-beta*kappa, beta*kappa, beta*kappa, -beta*kappa].
    Used to compute d f / d theta for the barrier control; the authors state that an accurate mapping is not required, but the sign is significant.
  • domain assumption The human input u has bounded energy over the time horizon.
    Required for the L(t) bound in the proof; physically, human forces are finite but not necessarily square-integrable over infinite time.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A control scheme for collaborative object transportation between a human and a quadruped robot using the MIGHTY suction cup." pith.science (2026). https://pith.science/paper/S6AI2VKT

@misc{pith2026250800584,
  author       = {Pith},
  title        = {Pith review of: A control scheme for collaborative object transportation between a human and a quadruped robot using the MIGHTY suction cup},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/S6AI2VKT}},
  note         = {Machine review of arXiv:2508.00584}
}
read the original abstract

In this work, a control scheme for human-robot collaborative object transportation is proposed, considering a quadruped robot equipped with the MIGHTY suction cup that serves both as a gripper for holding the object and a force/torque sensor. The proposed control scheme is based on the notion of admittance control, and incorporates a variable damping term aiming towards increasing the controllability of the human and, at the same time, decreasing her/his effort. Furthermore, to ensure that the object is not detached from the suction cup during the collaboration, an additional control signal is proposed, which is based on a barrier artificial potential. The proposed control scheme is proven to be passive and its performance is demonstrated through experimental evaluations conducted using the Unitree Go1 robot equipped with the MIGHTY suction cup.

Figures

Figures reproduced from arXiv: 2508.00584 by the authors.

Figure 1
Figure 1. Considered human-robot collaboration problem and setup. version of MIGHTY is employed ( [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. MIGHTY suction cup. The miniaturization of the MIGHTY system was made possible by its innovative rigid base, designed to accommo￾date a single electronics board. This board is fully equipped with all necessary pressure sensors, featuring embedded channels that directly link each chamber of the silicon cup to its respective pressure sensor. The silicon cup underwent itself some modifications, in order to enhance its … view at source ↗
Figure 3
Figure 3. MIGHTY suction cup sensor board. the z-axis of each sensor, i.e. to the i-th chamber, based on the values of pi-s and pv: fi = c(pi − hi(pv)), i = 1, ..., 4, (1) where hi(pv) = aipv + bi a linear regression function for removing the bias of the measurement, with ai , bi , c ∈ R parameters that are calibrated using ground truth measure￾ments. Remark 2. For calibrating ai , bi , c, we followed the follow￾ing rationale… view at source ↗
Figures from the paper (5 more)
Figure 5
Figure 5. Figure 5: Rationale behind (16). Remark 3. Notice that the gradient of the barrier artificial potential provided in (15) involves two control gains, namely k1 and k2; one for the non-linear and one for the linear term respectively. In order to tune those gains, one can arbitrari…
Figure 4
Figure 4. Figure 4: The barrier artificial potential and its derivative. Let us now proceed to the calculation of ∂W ∂θ of (9). Following the chain rule we get: ∂W ∂θ = ∂W ∂fm · ∂fm ∂fs · ∂fs ∂θ . (14) Calculating the partial derivatives analytically we get the following from (10): ∂W ∂fm…
Figure 8
Figure 8. Figure 8: Experiment involving the translation of the object which reflects the relatively high cognitive load rquired by the user. On the other hand, when using a relatively high damping, i.e. by setting ζ = ζ, the total energy transferred by the human to the robot is significa…
Figure 7
Figure 7. Figure 7: The gradient of the Barrier Artificial Potential (BAP) and the corresponding values of fi, i = 1, ..., 4 and fm in the experiment involving the rotation of the object. However, the range of this limitation can be broaden by utilizing faster measurements and/or a smalle…
Figure 9
Figure 9. Figure 9: Experiment involving the translation VI. CONCLUSIONS A control scheme for collaborative object transfer was proposed, between a human and a quadruped robot. The quadruped robot is equipped with the MIGHTY suction cup which is used both as a gripper and a force/torque s…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

31 extracted references · 31 canonical work pages

  1. [1]

    Human–robot collaboration for safe object transportation using force feedback,

    J. E. Solanes, L. Gracia, P. Mu ˜noz-Benavent, J. Valls Miro, M. G. Carmichael, and J. Tornero, “Human–robot collaboration for safe object transportation using force feedback,” Robotics and Autonomous Systems, vol. 107, pp. 196–208, 2018

    work page 2018

  2. [2]

    Safe and effective collaboration with a high-payload robot: A framework integrating novel hardware and software modules,

    A. Sidiropoulos, F. Dimeas, D. Papageorgiou, T. Prapavesis Semet- zidis, Z. Doulgeri, A. Zanella, F. Grella, K. Sagar, M. Jilich, A. Albini, G. Cannata, and M. Zoppi, “Safe and effective collaboration with a high-payload robot: A framework integrating novel hardware and software modules,” IEEE Robotics & Automation Magazine , pp. 2– 11, 2023

    work page 2023

  3. [3]

    A variable admittance controller for human-robot manipulation of large inertia objects,

    A. Sidiropoulos, T. Kastritsi, D. Papageorgiou, and Z. Doulgeri, “A variable admittance controller for human-robot manipulation of large inertia objects,” in 2021 30th IEEE International Conference on Robot & Human Interactive Communication (RO-MAN) , pp. 509–514, 2021

    work page 2021

  4. [4]

    Compliant human–robot object transfer based on modular 3-axis force sensor for collaborative manufacturing,

    H. Hua, Z. Liao, Y . Liu, X. Wu, J. Zhao, and J. Song, “Compliant human–robot object transfer based on modular 3-axis force sensor for collaborative manufacturing,”ISA Transactions, vol. 141, pp. 482–495, 2023

    work page 2023

  5. [5]

    A passive power-based control strategy for phri tasks with omni-directional robotic mobile platforms,

    T. Kastritsi and A. Ajoudani, “A passive power-based control strategy for phri tasks with omni-directional robotic mobile platforms,” IEEE Robotics and Automation Letters , vol. 9, no. 8, pp. 6959–6966, 2024

    work page 2024

  6. [6]

    Enhancing human–robot collaborative transportation through obstacle-aware vibrotactile warning and virtual fixtures,

    D. Sirintuna, T. Kastritsi, I. Ozdamar, J. M. Gandarias, and A. Ajoudani, “Enhancing human–robot collaborative transportation through obstacle-aware vibrotactile warning and virtual fixtures,” Robotics and Autonomous Systems , vol. 178, p. 104725, 2024

    work page 2024

  7. [7]

    Human-robot collaborative control for handling and transfer objects,

    G. P. Moreno, N. D. De la Cruz, J. S. Ortiz, and V . H. Andaluz, “Human-robot collaborative control for handling and transfer objects,” in Applied Technologies (M. Botto-Tobar, S. Montes Le ´on, O. Cama- cho, D. Ch ´avez, P. Torres-Carri ´on, and M. Zambrano Vizuete, eds.), (Cham), pp. 96–110, Springer International Publishing, 2021

    work page 2021

  8. [8]

    M. D. Kennedy III, Modeling and control for robotic assistants: Single and multi-robot manipulation. PhD thesis, University of Pennsylvania, 2019

    work page 2019

Show all 31 references
  1. [9]

    Mapping human inten- tions to robot motions via physical interaction through a jointly-held object,

    Y . Karayiannidis, C. Smith, and D. Kragic, “Mapping human inten- tions to robot motions via physical interaction through a jointly-held object,” in The 23rd IEEE International Symposium on Robot and Human Interactive Communication , pp. 391–397, 2014

    2014

  2. [10]

    D. J. Agravante, Human-humanoid collaborative object transportation. Theses, Universit´e Montpellier, Dec. 2015

    2015

  3. [11]

    Human-humanoid collaborative carrying,

    D. J. Agravante, A. Cherubini, A. Sherikov, P.-B. Wieber, and A. Kheddar, “Human-humanoid collaborative carrying,” IEEE Trans- actions on Robotics , vol. 35, no. 4, pp. 833–846, 2019

    2019

  4. [12]

    A new approach on hu- man–robot collaboration with humanoid robot rh-2,

    C. A. Monje, P. Pierro, and C. Balaguer, “A new approach on hu- man–robot collaboration with humanoid robot rh-2,” Robotica, vol. 29, no. 6, p. 949–957, 2011

    2011

  5. [13]

    Adaptive cooperative control for human-robot load manipulation,

    C. R. d. Cos and D. V . Dimarogonas, “Adaptive cooperative control for human-robot load manipulation,” IEEE Robotics and Automation Letters, vol. 7, no. 2, pp. 5623–5630, 2022

    2022

  6. [14]

    Human-robot collaborative object transfer using human motion prediction based on dynamic movement primitives,

    A. Sidiropoulos, Y . Karayiannidis, and Z. Doulgeri, “Human-robot collaborative object transfer using human motion prediction based on dynamic movement primitives,” in 2019 18th European Control Conference (ECC), pp. 2583–2588, 2019

    2019

  7. [15]

    Human-robot collaborative object transfer using human motion prediction based on cartesian pose dynamic movement primitives,

    A. Sidiropoulos, Y . Karayiannidis, and Z. Doulgeri, “Human-robot collaborative object transfer using human motion prediction based on cartesian pose dynamic movement primitives,” in 2021 IEEE Inter- national Conference on Robotics and Automation (ICRA) , pp. 3758– 3764, 2021

    2021

  8. [16]

    Anymal - a highly mobile and dynamic quadrupedal robot,

    M. Hutter, C. Gehring, D. Jud, A. Lauber, C. D. Bellicoso, V . Tsounis, J. Hwangbo, K. Bodie, P. Fankhauser, M. Bloesch, R. Diethelm, S. Bachmann, A. Melzer, and M. Hoepflinger, “Anymal - a highly mobile and dynamic quadrupedal robot,” in 2016 IEEE/RSJ Interna- tional Conferen...

    2016

  9. [17]

    Learning quadrupedal locomotion over challenging terrain,

    J. Lee, J. Hwangbo, L. Wellhausen, V . Koltun, and M. Hutter, “Learning quadrupedal locomotion over challenging terrain,” Science Robotics, vol. 5, no. 47, p. eabc5986, 2020

    2020

  10. [18]

    Control of dynamic gaits for a quadrupedal robot,

    C. Gehring, S. Coros, M. Hutter, M. Bloesch, M. A. Hoepflinger, and R. Siegwart, “Control of dynamic gaits for a quadrupedal robot,” in 2013 IEEE International Conference on Robotics and Automation , pp. 3287–3292, 2013

    2013

  11. [19]

    Contact planning for the anymal quadruped robot using an acyclic reachability- based planner,

    M. Geisert, T. Yates, A. Orgen, P. Fernbach, and I. Havoutis, “Contact planning for the anymal quadruped robot using an acyclic reachability- based planner,” in Towards Autonomous Robotic Systems (K. Althoe- fer, J. Konstantinova, and K. Zhang, eds.), (Cham), pp. 275–287, Spri...

    2019

  12. [20]

    Two-layer adaptive trajectory tracking controller for quadruped robots on slippery terrains,

    D.-E. Argiropoulos, D. Papageorgiou, M. Maravgakis, D. Drosakis, and P. Trahanias, “Two-layer adaptive trajectory tracking controller for quadruped robots on slippery terrains,” in 2023 IEEE-RAS 22nd International Conference on Humanoid Robots (Humanoids) , pp. 1–8, 2023

    2023

  13. [21]

    Probabilistic contact state estimation for legged robots using inertial information,

    M. Maravgakis, D.-E. Argiropoulos, S. Piperakis, and P. Trahanias, “Probabilistic contact state estimation for legged robots using inertial information,” in 2023 IEEE International Conference on Robotics and Automation (ICRA), pp. 12163–12169, 2023

    2023

  14. [22]

    Modeling and optimal control of rescue quadruped robot with high payload,

    N. Hu, S. Li, D. Huang, and F. Gao, “Modeling and optimal control of rescue quadruped robot with high payload,” in Mechanism and Ma- chine Science (X. Zhang, N. Wang, and Y . Huang, eds.), (Singapore), pp. 521–535, Springer Singapore, 2017

    2017

  15. [23]

    Development of quadruped robot for inspection of underground pipelines in nuclear power plants,

    Y . Jang, W. Seol, K. Lee, K.-S. Kim, and S. Kim, “Development of quadruped robot for inspection of underground pipelines in nuclear power plants,” Electronics Letters, vol. 58, no. 6, pp. 234–236, 2022

    2022

  16. [24]

    Vision aided dynamic exploration of unstructured terrain with a small-scale quadruped robot,

    D. Kim, D. Carballo, J. Di Carlo, B. Katz, G. Bledt, B. Lim, and S. Kim, “Vision aided dynamic exploration of unstructured terrain with a small-scale quadruped robot,” in 2020 IEEE International Conference on Robotics and Automation (ICRA) , pp. 2464–2470, 2020

    2020

  17. [25]

    Quadruped guidance robot for the visually impaired: A comfort-based approach,

    Y . Chen, Z. Xu, Z. Jian, G. Tang, L. Yang, A. Xiao, X. Wang, and B. Liang, “Quadruped guidance robot for the visually impaired: A comfort-based approach,” in 2023 IEEE International Conference on Robotics and Automation (ICRA) , pp. 12078–12084, 2023

    2023

  18. [26]

    Robotic guide dog: Leading a human with leash-guided hybrid physical in- teraction,

    A. Xiao, W. Tong, L. Yang, J. Zeng, Z. Li, and K. Sreenath, “Robotic guide dog: Leading a human with leash-guided hybrid physical in- teraction,” in 2021 IEEE International Conference on Robotics and Automation (ICRA), pp. 11470–11476, 2021

    2021

  19. [27]

    Gaze-guided semi-autonomous quadruped robot for enhanced assisted living,

    P. Joseph, D. Plozza, L. Pascarella, and M. Magno, “Gaze-guided semi-autonomous quadruped robot for enhanced assisted living,” in 2024 IEEE Sensors Applications Symposium (SAS) , pp. 1–6, 2024

    2024

  20. [28]

    Adaptive interactive control of human and quadruped robot load motion,

    S. Gu, F. Meng, B. Liu, X. Chen, Z. Yu, and Q. Huang, “Adaptive interactive control of human and quadruped robot load motion,” IEEE/ASME Transactions on Mechatronics , pp. 1–12, 2024

    2024

  21. [29]

    Mighty: Multi-functional suction cup for object gripping and surface attach- ment,

    E. Papadakis, M. Sigalas, M. Vangos, and P. Trahanias, “Mighty: Multi-functional suction cup for object gripping and surface attach- ment,” in 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , pp. 1–8, 2023

    2023

  22. [30]

    The gem- c controller for load compensation in object manipulation,

    E. Papadakis, M. Sigalas, M. Vangos, and P. Trahanias, “The gem- c controller for load compensation in object manipulation,” in 2024 IEEE International Conference on Robotics and Automation (ICRA) , pp. 13904–13909, 2024

    2024

  23. [31]

    On the stability of robot kinesthetic guidance in the presence of active constraints,

    T. Kastritsi, D. Papageorgiou, and Z. Doulgeri, “On the stability of robot kinesthetic guidance in the presence of active constraints,” in 2018 European Control Conference (ECC) , pp. 622–627, 2018

    2018

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.