Pith. sign in

REVIEW 3 major objections 6 minor 27 references

Picking by Tilting: In-Hand Manipulation for Object Picking using Effector with Curved Form

T0 review · 3 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read A curved passive palm, helped by two flat supports, can reliably pick objects too large for prehensile grasping by tilting them and tucking underneath.

desk verdict A credible proof-of-concept for picking large objects with a passive curved palm and a wall, but the planning model is not tied to the executed paths, so the mechanics section is scaffolding rather than the explanation of the successful trials. read the letter →

arxiv 2411.16055 v1 pith:PKMVL335 submitted 2024-11-25 cs.RO

Yanshu Song , Abdullah Nazir , Darwin Lau , Yun Hui Liu This is my paper · ORCID
classification cs.RO
keywords nonprehensilemanipulationin-handtiltingforce-closurepassiveend-effectorquasistaticobjectpickingcurvedpalm
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper tries to show that an object too large for any prehensile grasp can still be picked up reliably by a robot that uses only a curved, passive palm and two flat environmental surfaces. The key maneuver is to tilt the object against a wall-like support, creating a gap underneath, and then tuck the palm into that gap to support the object against gravity. If true, this would let simple hardware—just two motor axes or a conventional arm with a rigid curved attachment—perform a manipulation task usually thought to require a dexterous hand. The authors support the claim with a quasistatic planar analysis of tilting, a force-closure-based planning method, and picking experiments on objects of various sizes and shapes with two different robot setups.

What carries the argument

The central object is the quasistatic planar contact model with three frictional point contacts, analyzed through force-closure via composite wrench cones and moment labeling. The configuration space (θ, δ)—the object's tilt angle and the location of the object-palm contact along the object's edge—is precomputed with a linear-programming force-closure test, and tilting is planned as a straight-line path between an initial and a target configuration lying inside the force-closure region. At the target configuration, first-order (Reuleaux) mobility analysis determines whether the palm's orientation kinematically prevents the object from ungrasping by clockwise rotation.

What would settle it

Run the tilt with a palm of known curvature while tracking the object's pose and contact point; if the measured (θ, δ) trajectory leaves the precomputed force-closure region before δ reaches zero, the object will fall exactly as in Fig. 7(a), disproving the sufficiency of the straight-line planning assumption.

Watch

Extended reading notes

Core claim

The paper claims that the dexterous interactions needed to pick an object too large for prehensile grasping—tilting it up against a wall-like support and then tucking a curved passive palm underneath—can be realized with a zero-DOF rigid end-effector and simple control. It formulates picking as a quasistatic planar process with three frictional point contacts (object with bottom support, side support, and palm), analyzes two feasible tilting modes (two-contact wedged rotation and three-contact sliding), and plans a tilting path as a straight line in the (θ, δ) configuration space that stays inside the force-closure region. Experiments with a custom two-DOF palm and a conventional six-DOF arm demonstrate reliable picking of several objects, with reorienting the side support to a more acute angle (61°) resolving a failure mode where the object falls during tilting.

Load-bearing premise

The plan assumes that as the object tilts, the contact point on the palm moves monotonically so that δ decreases while θ increases, and that a straight-line path in the (θ, δ) space stays inside the force-closure region; the authors admit that the actual open-loop path deviates because the palm's curvature is not modeled.

Editorial extensions

If this is right

  • Objects too large for parallel-jaw or multi-finger grips can be picked with a single curved rigid palm, using environmental surfaces as the other 'fingers.'
  • The technique works without force/torque sensors: a motion-force hybrid control (position on one motor, constant torque on the other) suffices for reliable tilting with fewer commands.
  • Reorienting the side support to a more acute angle redirects the side contact normal downward and suppresses the failure mode where the object falls before the palm tucks underneath.
  • High-friction palm material enlarges the set of configurations in which the object is force-closure during tilting, compensating for open-loop execution.
  • The planar quasistatic analysis extends to extruded or revolved inverted truncated cone objects, giving a class of 3D shapes for which the method is applicable.

Reading between the lines

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

  • If the rolling assumption is the bottleneck, a closed-loop controller using tactile or proprioceptive feedback to track (θ, δ) could convert the open-loop blueprint into a robust plan without changing the hardware.
  • The same tilt-and-tuck template may generalize to other environment layouts, such as a wall corner and a ceiling, or to compliant objects where local contact curvature stabilizes tilting, as hinted by the cardboard box experiments.
  • The authors' comparison of support angles suggests a design principle for fixtures: making the blocking surface more acute turns an unreliable three-contact slide into a repeatable wedge-and-tilt maneuver.
  • A testable extension is to precompute the force-closure region for a library of palm profiles and choose a palm shape that maximizes the (θ, δ) area where tilting succeeds, since palm geometry is currently ignored.
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

3 major / 6 minor

Summary. The paper proposes a nonprehensile picking method in which a curved, passive end-effector (the 'palm') tilts an object against two flat supports and then tucks underneath it to obtain a gravity grasp. The authors formulate the task as a quasistatic planar problem with three frictional contacts, classify two feasible tilting modes and one infeasible mode, and compute force-closure regions in a (theta, delta) configuration space to plan a tilting path. They demonstrate the method on a custom two-DOF palm and on a UR3 arm with several objects, reporting mostly 5/5 success rates, with one 0/5 condition for a steel box at a 90-degree support angle that was fixed by changing the support angle to 61 degrees. The paper concludes that simple hardware and control can realize the dexterous interactions needed for this form of picking.

Significance. If the result holds, the paper provides a useful addition to nonprehensile and extrinsic-dexterity manipulation: a passive curved palm, with no additional degrees of freedom or force/torque sensing, can pick objects too large for prehensile grasping by exploiting environment contacts. The mechanics analysis in Sec. IV gives a clear qualitative classification of tilting modes, and the force-closure plots in Fig. 3(b) offer a falsifiable prediction about which configurations should be stable. The hardware demonstration, including a conventional robot-arm implementation, is a meaningful proof of concept. However, the paper does not release code or raw experimental data, and the connection between the theoretical planning model and the open-loop experiments is not established, which limits the strength of the claimed reliability result.

major comments (3)
  1. [Sec. IV-B and Sec. V-D] The planning strategy in Sec. IV-B assumes that during tilting the object rolls on the curved palm so that delta decreases monotonically as theta increases, and that a straight-line path in (theta, delta) between initial and target configurations lies inside the force-closure region. Neither property is derived from the palm profile or from the contact model. Sec. V-D then explicitly concedes that the actual open-loop path differs from the nominally feasible path because the geometry of the palm is ignored. Since no executed (theta, delta) trajectories are reported, the experiments do not show that the object remained inside the force-closure region during tilting. This gap is load-bearing: without it, the theoretical analysis in Sec. IV-B does not explain the successful trials. The paper should either measure or estimate the executed paths, or explicitly reposition the planning analysis as a heuristic and support the empirical claim independently.
  2. [Table I and Sec. V-C] The experimental evidence for reliable picking is thin: only 5 trials per condition are reported, the steel box at 90 degrees failed 0/5, and the subsequent fixture change to 61 degrees is explained only qualitatively via Sec. IV-A. The text says that reorienting support #2 redirects the contact normal downward, which predicts the direction of the improvement, but it does not provide a quantitative derivation of the 61-degree threshold, a parameter sweep, or trials across a broader range of support angles and object shapes. Given the small sample sizes, the success-rate differences between conditions (e.g., 3/5 versus 5/5) are not statistically meaningful, and the teleoperated execution makes it difficult to separate the method's robustness from operator skill. Please report confidence intervals, additional trials, or a clear statement that the empirical claim is at the level of a demonstration rather than a reliability characterization.
  3. [Sec. IV-B, experimental usage] After presenting the force-closure planning method, the paper states that 'feasible initial and target configurations are chosen empirically' in the experiments. This creates a disconnect between the proposed planning blueprint and the actual experimental protocol. If the planning method was not used to select configurations, then the experiments do not validate the planning method; they only validate a manually chosen implementation of the general idea. The authors should clarify whether the force-closure computation was used at all in choosing the experimental trajectories, and if not, what role the planning section is intended to play in the claimed contribution.
minor comments (6)
  1. [Sec. V-A] The word 'ellipsical' should be 'elliptical' in the description of the palm's curved surface.
  2. [Sec. V-B] The phrase 'gently slammed' in the description of the initial contact is contradictory; consider replacing it with 'gently pressed' or 'brought into contact'.
  3. [Sec. IV-B] The sentence 'Tilting supposedly begins (ends) when theta = 0 (delta = 0)' is vague; please define the initial and target configurations precisely, including the relation of these values to the object and palm geometry.
  4. [Fig. 3(b)] The caption and text do not state the object dimensions or the exact friction coefficients used to generate the force-closure regions; adding these details would make the plot reproducible.
  5. [Sec. V-D] The claim that high-friction rubber on the palm 'seems sufficient' to attain high success rates is not tested by any ablation; comparing a lower-friction palm or a quantitative friction sensitivity study would strengthen this assertion.
  6. [Reproducibility] The paper does not provide the LP force-closure software, experimental data, or trial logs; releasing these would substantially improve the reproducibility of the claimed success rates.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the tilting mechanics and force-closure planning are derived from standard contact/wrench-cone models and are not fitted to or defined by the experimental successes.

full rationale

The paper's derivation chain is self-contained with respect to its claimed inputs. The mechanics of Sec. IV-A derives feasible tilting modes from rigid-body friction cones and moment-label/wrench-cone analysis, citing standard references (Mason's Mechanics of Robotic Manipulation and Lynch and Park's Modern Robotics). The force-closure regions in Fig. 3(b) are computed by a linear-programming test over sampled (theta, delta) configurations from the stated contact model and friction coefficients, not from the experiment outcomes. The planning step in Sec. IV-B constructs a straight-line path inside the computed force-closure region; this is a genuine prediction from the model rather than a restatement of the experiment results. The experiments in Sec. V are then reported as separate evidence. The acknowledged limitation in Sec. V-D that the actual open-loop path differs from the nominal feasible path because palm geometry is ignored is a validation gap and a correctness risk, but it is not circularity: the model was not fitted to the successes, and no fitted parameter is renamed as a prediction. The only overlap with prior work by an author, reference [6] by co-author Nazir, appears in the related-work survey and is not load-bearing for any derivation. No self-definitional step, fitted-input-called-prediction step, or uniqueness/ansatz-importing citation chain was found. The central claim that a curved passive palm can realize the required dexterous interactions is supported by the experiments and by the independent mechanics analysis, even though the planning model is incomplete. Therefore the circularity score is 0.

Assumptions & free parameters 4 free parameters · 6 assumptions · 0 invented entities

No new physical entities are introduced; free parameters and assumptions are limited to contact friction values, support angle, and empirically chosen configurations, plus the unproven rolling and straight-line path assumptions in the planning section.

free parameters (4)
  • Friction coefficient at palm contact mu_C = 0.1 and 0.2 in planning; high-friction rubber in experiments, value not measured
    Chosen for the force-closure plots in Fig. 3(b), not measured. The qualitative claim that higher mu_C enlarges the secure region is robust, but the plotted boundary depends on this assumed value.
  • Friction coefficients at supports mu_A and mu_B = 0.1 in planning
    Assumed in the force-closure computation; no friction measurement for support surfaces is reported.
  • Initial and target tilting configurations (theta, delta) = chosen empirically per experiment
    Sec. IV-B states feasible initial and target configurations are chosen empirically to facilitate picking; these choices are not derived from the mechanics model.
  • Angle between the two support surfaces = 90 degrees in planning; 61 degrees for the successful steel-box trials
    The support angle is a design parameter that strongly affects success: steel boxes fail at 90 degrees and succeed at 61 degrees, so the reliability claim is conditional on this hand-set fixture choice.
assumptions (6)
  • standard math Rigid body mechanics with Coulomb friction and frictional point contacts apply to the object-environment interaction.
    Used throughout Sec. IV for friction cones, force-closure, and moment labeling; not verified for the deformable cardboard object.
  • domain assumption The manipulation is quasistatic and planar, in a plane normal to the two supports.
    Sec. III states the task is modeled as a quasistatic process in a 2D plane; out-of-plane motion, dynamics, and inertia are neglected.
  • domain assumption The target object is a trapezoid whose top edge is not shorter than its bottom edge, extruded or revolved into an inverted truncated cone.
    Sec. III defines this object class; the experiments use boxes and a beverage container that approximate it.
  • ad hoc to paper During tilting, the object rolls on the curved palm such that delta decreases monotonically as theta increases.
    Sec. IV-B states this as an expectation; it is not derived and it supports the straight-line planning path.
  • ad hoc to paper A straight-line path in (theta, delta) between selected configurations remains inside the force-closure region.
    Sec. IV-B proposes such a path; Sec. V-D admits the actual open-loop path differs, so the region containment is assumed rather than guaranteed.
  • ad hoc to paper The geometry of the palm can be ignored when planning; only the contact point C and the palm orientation matter.
    Sec. V-D says modeling the contact evolution will necessitate taking into consideration the geometry of the palm, which is ignored in the current work. This is load-bearing for the planned force-closure regions.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Picking by Tilting: In-Hand Manipulation for Object Picking using Effector with Curved Form." pith.science (2026). https://pith.science/paper/PKMVL335

@misc{pith2026241116055,
  author       = {Pith},
  title        = {Pith review of: Picking by Tilting: In-Hand Manipulation for Object Picking using Effector with Curved Form},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PKMVL335}},
  note         = {Machine review of arXiv:2411.16055}
}
read the original abstract

This paper presents a robotic in-hand manipulation technique that can be applied to pick an object too large to grasp in a prehensile manner, by taking advantage of its contact interactions with a curved, passive end-effector, and two flat support surfaces. First, the object is tilted up while being held between the end-effector and the supports. Then, the end-effector is tucked into the gap underneath the object, which is formed by tilting, in order to obtain a grasp against gravity. In this paper, we first examine the mechanics of tilting to understand the different ways in which the object can be initially tilted. We then present a strategy to tilt up the object in a secure manner. Finally, we demonstrate successful picking of objects of various size and geometry using our technique through a set of experiments performed with a custom-made robotic device and a conventional robot arm. Our experiment results show that object picking can be performed reliably with our method using simple hardware and control, and when possible, with appropriate fixture design.

Figures

Figures reproduced from arXiv: 2411.16055 by the authors.

Figure 1
Figure 1. Sequence of snapshots (clockwise from the top-left) showing our [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Feasibility of initial tilting with two contacts (a) and three contacts (b-c). In (a), the object is wedged between contacts [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. (a) Real tilting scenario (left) modeled in a plane normal to the two [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: With an appropriately oriented palm at the target configuration, [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]
Figure 6
Figure 6. Figure 6: Teleoperated object picking with our robotic palm. (a) Snapshots of experiment with a yoga block (top row) and a beverage container (bottom [PITH_FULL_IMAGE:figures/full_fig_p005_6.png]
Figure 7
Figure 7. Figure 7: Object picking with a conventional manipulator. Snapshots of (a) unsuccessful, and (b) successful picking experiments. [PITH_FULL_IMAGE:figures/full_fig_p006_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

27 extracted references · 25 canonical work pages

  1. [1]

    Hands for dexterous manipulation and robust grasping: a difficult road toward simplicity,

    A. Bicchi, “Hands for dexterous manipulation and robust grasping: a difficult road toward simplicity,” IEEE Transactions on Robotics and Automation, vol. 16, no. 6, pp. 652–662, Dec 2000

    work page 2000

  2. [2]

    Robot hands and the mechanics of manipulation,

    M. T. Mason and J. K. Salisbury Jr, “Robot hands and the mechanics of manipulation,” 1985

    work page 1985

  3. [3]

    Nonprehensile palmar manipulation with a mobile robot,

    W. Huang and G. Holden, “Nonprehensile palmar manipulation with a mobile robot,” in Proceedings 2001 IEEE/RSJ International Confer- ence on Intelligent Robots and Systems. Expanding the Societal Role of Robotics in the the Next Millennium (Cat. No.01CH37180) , vol. 1, Oct 2001, pp. 114–119 vol.1

    work page 2001

  4. [4]

    Object picking through in-hand manipulation using passive end-effectors with zero mobility,

    C. Mucchiani, M. Kennedy, M. Yim, and J. Seo, “Object picking through in-hand manipulation using passive end-effectors with zero mobility,” IEEE Robotics and Automation Letters , vol. 3, no. 2, pp. 1096–1103, 2018

    work page 2018

  5. [5]

    Robotic edge- rolling manipulation: A grasp planning approach,

    A. Specian, C. Mucchiani, M. Yim, and J. Seo, “Robotic edge- rolling manipulation: A grasp planning approach,” IEEE Robotics and Automation Letters, vol. 3, no. 4, pp. 3137–3144, Oct 2018

    work page 2018

  6. [6]

    Rock-and-walk manipulation: Object locomotion by passive rolling dynamics and periodic active control,

    A. Nazir, P. Xu, and J. Seo, “Rock-and-walk manipulation: Object locomotion by passive rolling dynamics and periodic active control,” IEEE Transactions on Robotics , vol. 38, no. 4, pp. 2354–2369, Aug 2022

    work page 2022

  7. [7]

    An exploration of nonprehensile two-palm manipulation,

    M. Erdmann, “An exploration of nonprehensile two-palm manipulation,” The International Journal of Robotics Research , vol. 17, no. 5, pp. 485–503, 1998. [Online]. Available: https://doi.org/10.1177/027836499801700502

    work page doi:10.1177/027836499801700502 1998

  8. [8]

    Tactile dexterity: Manipulation primitives with tactile feedback,

    F. R. Hogan, J. Ballester, S. Dong, and A. Rodriguez, “Tactile dexterity: Manipulation primitives with tactile feedback,” in 2020 IEEE International Conference on Robotics and Automation (ICRA) , May 2020, pp. 8863–8869

    work page 2020

Show all 27 references
  1. [9]

    Exploitation of environmental constraints in human and robotic grasping,

    C. Eppner, R. Deimel, J. ´Alvarez Ruiz, M. Maertens, and O. Brock, “Exploitation of environmental constraints in human and robotic grasping,” The International Journal of Robotics Research, vol. 34, no. 7, pp. 1021–1038, 2015. [Online]. Available: https://doi.org/10.1177/02783...

    2015 doi

  2. [10]

    Shallow-depth insertion: Peg in shallow hole through robotic in-hand manipulation,

    C. H. Kim and J. Seo, “Shallow-depth insertion: Peg in shallow hole through robotic in-hand manipulation,” IEEE Robotics and Automation Letters, vol. 4, no. 2, pp. 383–390, April 2019

    2019

  3. [11]

    Extrinsic dexterity: In-hand manipulation with external forces,

    N. C. Dafle, A. Rodriguez, R. Paolini, B. Tang, S. S. Srinivasa, M. Erdmann, M. T. Mason, I. Lundberg, H. Staab, and T. Fuhlbrigge, “Extrinsic dexterity: In-hand manipulation with external forces,” in 2014 IEEE International Conference on Robotics and Automation (ICRA), May 20...

    2014

  4. [12]

    Planning for dexterous manipulation with sliding contacts,

    J. Trinkle and R. Paul, “Planning for dexterous manipulation with sliding contacts,” The International Journal of Robotics Research, vol. 9, no. 3, pp. 24–48, 1990. [Online]. Available: https://doi.org/10.1177/027836499000900302

    1990 doi

  5. [13]

    Dextrous manipulation with rolling contacts,

    L. Han, Y . Guan, Z. Li, Q. Shi, and J. Trinkle, “Dextrous manipulation with rolling contacts,” in Proceedings of International Conference on Robotics and Automation , vol. 2, April 1997, pp. 992–997 vol.2

    1997

  6. [14]

    Dexterous grippers: Putting nonholonomy to work for fine manipulation,

    A. Bicchi and A. Marigo, “Dexterous grippers: Putting nonholonomy to work for fine manipulation,” The International Journal of Robotics Research, vol. 21, no. 5-6, pp. 427–442, 2002. [Online]. Available: https://doi.org/10.1177/027836402321261968

    2002 doi

  7. [15]

    Design of a roller-based dexterous hand for object grasping and within-hand manipulation,

    S. Yuan, A. D. Epps, J. B. Nowak, and J. K. Salisbury, “Design of a roller-based dexterous hand for object grasping and within-hand manipulation,” in 2020 IEEE International Conference on Robotics and Automation (ICRA) , May 2020, pp. 8870–8876

    2020

  8. [16]

    Dynamic in-hand sliding manipulation,

    J. Shi, J. Z. Woodruff, P. B. Umbanhowar, and K. M. Lynch, “Dynamic in-hand sliding manipulation,” IEEE Transactions on Robotics, vol. 33, no. 4, pp. 778–795, Aug 2017

    2017

  9. [17]

    Progress in nonprehensile manipulation,

    M. T. Mason, “Progress in nonprehensile manipulation,” International Journal of Robotics Research, vol. 18, no. 11, pp. 1129–1141, Novem- ber 1999

    1999

  10. [18]

    The mechanics of fine manipulation by pushing,

    K. M. Lynch, “The mechanics of fine manipulation by pushing,” in Proceedings 1992 IEEE International Conference on Robotics and Automation, May 1992, pp. 2269–2276 vol.3

    1992

  11. [19]

    Pivoting: A new method of graspless manipulation of object by robot fingers,

    Y . Aiyama, M. Inaba, and H. Inoue, “Pivoting: A new method of graspless manipulation of object by robot fingers,” in Intelligent Robots and Systems ’93, IROS ’93. Proceedings of the 1993 IEEE/RSJ International Conference on , vol. 1, Jul 1993, pp. 136–143 vol.1

    1993

  12. [20]

    Toppling manipulation,

    K. M. Lynch, “Toppling manipulation,” in Robotics and Automation,

  13. [21]

    Robust execution of contact-rich motion plans by hybrid force-velocity control,

    Y . Hou and M. T. Mason, “Robust execution of contact-rich motion plans by hybrid force-velocity control,” in 2019 International Confer- ence on Robotics and Automation (ICRA) , May 2019, pp. 1933–1939

    2019

  14. [22]

    Picking thin objects by tilt-and-pivot manipulation and its application to bin picking,

    Z. Tong, T. He, C. H. Kim, Y . Hin Ng, Q. Xu, and J. Seo, “Picking thin objects by tilt-and-pivot manipulation and its application to bin picking,” in 2020 IEEE International Conference on Robotics and Automation (ICRA) , May 2020, pp. 9932–9938

    2020

  15. [23]

    M. T. Mason , Mechanics of Robotic Manipulation . Cambridge, MA: MIT Press, August 2001

    2001

  16. [24]

    K. M. Lynch and F. C. Park, Modern Robotics. Cambridge University Press, 2017

    2017

  17. [25]

    Planning for dexterous ungrasping: Secure ungrasping through dexterous manipulation,

    C. H. Kim, K. H. Mak, and J. Seo, “Planning for dexterous ungrasping: Secure ungrasping through dexterous manipulation,” IEEE Robotics and Automation Letters , vol. 7, no. 2, pp. 2234–2241, April 2022

    2022

  18. [26]

    Asada and K

    H. Asada and K. Youcef-Toumi, Direct-Drive Robots: Theory and Practice . The MIT Press, 06 1987. [Online]. Available: https://doi.org/10.7551/mitpress/2438.001.0001

    1987 doi

  19. [1999]

    1999 IEEE International Conference on , vol

    Proceedings. 1999 IEEE International Conference on , vol. 4. IEEE, 1999, pp. 2551–2557

    1999

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.