Pith. sign in

REVIEW 4 major objections 5 minor 14 references

Real-time Two-tape Control System in Vine robots

T0 review · 4 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

Pith's one-line read This paper shows a vine-style growing robot can be steered in real time by inducing surface wrinkles and fixing them with tape, producing repeatable ~21-degree turns.

desk verdict Real-time tip-applied tape steering is a genuine mechanical contribution with a credible single-turn result, but the paper's 'arbitrary many turns' and Dubins claims are unsupported and its angle model has an unexplained parameter shift. read the letter →

arxiv 2505.00969 v1 pith:IHTTQX76 submitted 2025-05-02 cs.RO

classification cs.RO
keywords vinerobotsoftroboticsreal-timesteeringsurfacewrinklesadhesivetapefixed-angleturnsDubinspathplanninggrowing
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 claims that a vine-like growing soft robot can be steered in real time by mechanically creating a fold, or wrinkle, in the uninflated tube at the growth point and fixing that fold in place with adhesive tape. The design feeds material from outside the inflated tube so wrinkles can be induced before inflation, and a compensation mechanism makes the fold asymmetric so the tube bends consistently when it inflates. The model predicts each wrinkle turns the robot about 21 degrees on a flat surface, and experiments report 21.5 ± 1.5 degrees for single turns and 21 ± 2 degrees for alternating turns. A discrete shortest-path planner turns the fixed-angle constraint into a sequence of left/right 21-degree increments, so the robot can follow planar paths without internal rigid actuators.

What carries the argument

The load-bearing mechanism is the wrinkle inducer and compensation system: two support arms lift the uninflated tube to create a fold while a lower bonding arm presses one of two exterior tape ribbons onto the fold, and a compensation arm first adheres the opposite ribbon so that side unfolds on inflation. The central identity is θ′ = (Dπ/L) cos(45°), the planar projection of the wrinkle-induced bend, where D is wrinkle length, L is flattened tube width, and 45° is the tape's quarter-point position around the circumference. This identity converts a fixed mechanical stroke into a repeatable turning angle and lets the controller treat each wrinkle as a discrete 21-degree turn.

What would settle it

Measure the wrinkle fold length and tape lateral position on every wrinkle of a ten-turn sequence and compare each measured turn angle with θ′ = (Dπ/L) cos(45°); if the angle drifts while D and tape alignment stay fixed, the geometric model is wrong, and if D or tape alignment drifts first, the fixed-angle claim fails.

Watch

Extended reading notes

Core claim

The central claim is that a Vine robot can be given arbitrary many planar turns in real time by inducing surface wrinkles at its growing end and taping them, without sacrificing the robot's soft body. Each wrinkle shortens one side of the tube by twice the wrinkle length, producing a turning angle θ = D/r = Dπ/L, which for D = 19 mm and L = 105 mm is 30.3 degrees in 3D; because the tape is placed at quarter-point positions, the planar projection is θ′ = θ cos(45°) ≈ 21.44°. Experiments confirm this: first-turn angle 21.5 ± 1.5°, alternating left/right sequences average 21 ± 2°, though ten consecutive same-direction turns degrade as material shifts and air leaks. The claim is that this mechanism is the first to combine real-time steering, multi-directional turns, and structural softness in a growing vine robot, with path planning reduced to discrete 21-degree steps.

Load-bearing premise

The 21-degree result depends on the wrinkle length staying at 19 mm and the tape staying bonded at the quarter-point positions, and the paper's own error analysis shows these are precisely what drift after repeated turns.

Editorial extensions

If this is right

  • Because each wrinkle yields a repeatable fixed-angle turn, the robot can execute arbitrary planar trajectories as sequences of left/right 21-degree increments, with a shortest-path planner choosing the sequence.
  • The external material feed lets the wrinkle be formed before inflation, so turning control happens in real time at the growth point rather than being fixed before deployment.
  • Alternating left and right turns stay within 2 degrees for at least ten turns, so bidirectional steering is usable immediately; the robot remains a soft continuum structure throughout.
  • The fixed-angle limit is a design choice, not a physical ceiling: varying the wrinkle length would produce arbitrary angles, extending the same mechanism to general paths.

Reading between the lines

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

  • The cos(45°) projection suggests a tunable steering law: if the tape's circumferential position could be varied, the same wrinkle mechanism would produce a continuous family of planar angles between 0° and 30.3°.
  • The degradation after consecutive same-direction turns points to a control fix the paper does not test: alternating turn directions, or periodically re-centering the tape, may act as an error-correcting strategy and extend the usable turn count.
  • The compensation principle—pre-adhering the opposite side so a symmetric fold unfolds asymmetrically—could transfer to other soft growing robots as a way to create programmed bends without internal actuators.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper presents a vine robot steering system that creates controlled turns in real time by using an end-device to induce surface wrinkles in the growing tube and fix them with adhesive tape. The authors derive a geometric model that predicts a fixed turning angle of about 21 degrees, describe a modified Dubins path planner intended to convert sequences of these discrete turns into a navigable path, and report experiments measuring single-turn, consecutive same-direction, and alternating left-right turn performance. The central hardware claim is that single turns are repeatable at 21.5 degrees plus or minus 1.5 degrees, with alternating sequences averaging 21 degrees plus or minus 2 degrees over the first ten turns.

Significance. If the single-turn result is robust, the contribution is a genuinely simple, real-time steering mechanism for a soft growing robot that preserves the soft body and does not require pre-shaped paths or rigid actuators. The external material feed and compensation mechanism are a plausible engineering advance, and the paper provides a useful geometric model for estimating the turn angle from wrinkle geometry, subject to the concerns below. However, the broader claims in the abstract and conclusion about 'arbitrary many turns' and a working Dubins path planner are not substantiated by the experiments, and the available data indicate a hard limit of roughly ten consecutive same-direction turns. The paper would be significantly strengthened by reframing these claims and by providing the missing experimental and algorithmic details.

major comments (4)
  1. [Section IV, Eq. (1)] The model uses D=19 mm in Eq. (1), but Section III-A states that the wrinkle inducer arm's 15-degree rotation produces a precise 18 mm fold. This discrepancy is unexplained. Because the predicted angle scales linearly with D, the agreement between the model's 21.44 degrees and the experimental 21.5 degrees depends on the choice of 19 mm. Please justify the 19 mm value (for example, by reporting a direct measurement of the fold length) or redo the prediction with the design value D=18 mm, which yields theta' = 20.3 degrees and still lies within the reported experimental uncertainty. As written, the model appears to use an adjusted parameter rather than a first-principles value.
  2. [Abstract, Section V, Section VI-A] The abstract's claim of 'arbitrary many turns using a Dubins path planner' is contradicted by the paper's own results. Algorithm 1 is never executed in any experiment in Section VI, and Section VI-A reports that beyond ten consecutive same-direction wrinkles the bend angle systematically decreases and the system fails. This directly refutes the 'arbitrary many turns' wording. The abstract and conclusion should be revised to state that the planner is proposed but not experimentally validated, and that sustained same-direction turning is currently limited to about ten wrinkles.
  3. [Section V, Algorithm 1] The pseudocode of Algorithm 1 does not implement the behavior described in the text. The text says that a 63-degree turn uses three consecutive 21-degree turns, but the algorithm as written searches for a single intersection point between discretized rays and returns a path length, not a sequence of discrete turn commands. There is no loop that would generate a series of 21-degree waypoints. Please provide the actual planning algorithm that generates discrete-turn sequences, or clearly label the current pseudocode as a conceptual sketch that has not been implemented.
  4. [Section VI] The experimental section does not report a measurement protocol or the number of trials underlying the statistics in Fig. 7 and the text. There is no description of how the turning angle was measured (e.g., protractor, image processing, tracking markers), how many runs were performed for each scenario, or how the error bars are defined. Without this information, the claims of '21.5 plus or minus 1.5 degrees' and '21 plus or minus 2 degrees' cannot be independently assessed, and the paper's key quantitative result lacks adequate support. Please add a clear experimental methodology subsection.
minor comments (5)
  1. [Section III-A] The text states that the fold length is determined by the length of the support arms and later says that a 15-degree arm rotation produces an 18 mm fold. Please clarify which geometric quantity (arm length or rotation angle) is the controlled parameter and how the 18 mm value is obtained.
  2. [Section IV] The sentence 'the outer arc of the wrinkle maintains this 2D length' is difficult to follow. Please expand the geometric derivation so that the factor of 2D in the length reduction, and hence theta = D/r, is clearly explained.
  3. [Section VI-B] The error analysis says that tape misalignment causes angular errors despite maintaining consistent wrinkle lengths, but this seems to imply that D remains constant while the angle changes. Please explain how D can stay fixed when the tape position relative to the wrinkle drifts, given that the model in Eq. (1) assumes D is the sole geometric determinant of the angle.
  4. [Fig. 5] The caption for Fig. 5 shows patterns R, RRR, and RLRLL but no quantitative angles or path measurements for these multi-turn demonstrations. Reporting the measured angles for each turn in these sequences would strengthen the evidence for multi-turn capability.
  5. [References] The URL in reference [13] appears to contain spacing and possible transcription errors. Please verify that the link is complete and functional.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the geometric angle model uses measured wrinkle length as an input and is confirmed by experiment; the unvalidated planner claim is a validation gap, not a circular derivation.

full rationale

The paper's central derivation (Section IV) is a direct geometric model: θ = Dπ/L, with D the wrinkle length and L the tube width, followed by a cos(45°) projection for planar motion. D is a physical input (the fold length produced by the wrinkle inducer), not the output angle; the model does not define D in terms of θ, and no fitting step is described. The predicted 21.44° is consistent with the measured 21.5°±1.5° but is not equal to it by construction — with the design value D=18 mm the formula gives about 21.8°, also within the error bar, showing the result is not a knife-edge fit. The apparent mismatch between the stated design fold (18 mm) and the modeling value (D=19 mm), and the arithmetic slip yielding θ=30.3° from D=19 mm, are presentation/consistency errors, not evidence that the prediction reduces to its input. The Dubins-planner extension (Section V) is indeed never executed in the experiments, and the same-direction multi-turn data show degradation; however, that is an unsupported-claim / validation issue, not circular reasoning. No load-bearing self-citation or imported uniqueness theorem appears. Therefore no significant circularity is present.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

The paper's central prediction is a one-parameter geometric model: with L=105 mm and D=19 mm it yields 21.44 degrees. D is the only adjustable quantity, but it is not independently fixed by the design specs (18 mm vs 19 mm). The rest of the ledger consists of domain assumptions about material inextensibility, the cos(45-degree) projection, and the compensation mechanism, all of which the experiments show can fail under repeated use.

free parameters (1)
  • Wrinkle length D = 19 mm
    The angle formula theta = D*pi/L is linear in D. The mechanical section states the arm rotation produces an 18 mm fold, but the model uses 19 mm with no explanation, so D acts as an effective calibration constant for the reported 21-degree turn.
assumptions (4)
  • standard math Inflating the flattened tube produces a circular cross-section with radius r = L/pi.
    Used implicitly in Eq. (1): theta = D/r = D*pi/L requires the flattened width L to equal half the inflated circumference.
  • domain assumption The PVC material does not stretch, so the wrinkle shortens the compressed side by 2D and the outer side conserves length.
    Section IV states 'the PVC tube material has very little stretch, the length is conserved' to justify the geometry.
  • domain assumption The 3D turn projects onto the table as theta cos(45 degrees) because the tape sits at quarter-point positions.
    Section IV: 'the actual angle on the table should be multiplied by cos(45 degrees).' This projection is asserted, not derived.
  • domain assumption The compensation mechanism keeps the non-wrinkle side unfolded and the feed material aligned with the rollers.
    The entire wrinkle asymmetry depends on this; Sections VI-B and VII-B-C show that alignment and leakage failures occur in practice.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Real-time Two-tape Control System in Vine robots." pith.science (2026). https://pith.science/paper/IHTTQX76

@misc{pith2026250500969,
  author       = {Pith},
  title        = {Pith review of: Real-time Two-tape Control System in Vine robots},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IHTTQX76}},
  note         = {Machine review of arXiv:2505.00969}
}
read the original abstract

This paper focuses on how to make a growing Vine robot steer in different directions with a novel approach to real-time steering control by autonomously applying adhesive tape to induce a surface wrinkles. This enabling real-time directional control with arbitrary many turns while maintaining the robot's soft structure. This system feeds growing material external to the tube. The design achieves fixed-angle turns in 2D space. Through experimental validation, we demonstrate repeated 21-degree turns using a Dubins path planner with minimal error, establishing a foundation for more versatile Vine robot applications. This approach combines real-time control, multi-degree-of-freedom steering, and structural flexibility, addressing key challenges in soft robotics.

Figures

Figures reproduced from arXiv: 2505.00969 by the authors.

Figure 1
Figure 1. (a) Overview of robot. (b) CAD model of end-device. [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. (a)-(c): Wrinkle Inducer in a neutral configuration prior to activation (d)-(f): Two arms on each side of the wrinkle [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 4
Figure 4. Wrinkle geometric parameters B. Workflow of Generating Wrinkles [PITH_FULL_IMAGE:figures/full_fig_p003_4.png] view at source ↗
Figures from the paper (5 more)
Figure 3
Figure 3. Figure 3: Compensation Mechanism [PITH_FULL_IMAGE:figures/full_fig_p003_3.png]
Figure 6
Figure 6. Figure 6: Segmented Dubins Path Algorithm 1 Generate Dubins Path procedure GENERATEDUBINSPATH(map, startPos, star￾tAngle, endPos, endAngle, pathType) map ← Precompute polygons and increments n ← 360/increment for i ← 1 to n do for j ← 1 to n do intP t ← INTERSECTION(map.startRay…
Figure 5
Figure 5. Figure 5: Experimental Validation of Multi-Turn Capabilities: (a) Single Right Turn (R) Demonstration, (b) Triple Consecutive [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 7
Figure 7. Figure 7: Performance Evaluation. Group 1 tests single turns. Group 2 tests sequential turns in one direction. Group 3 tests [PITH_FULL_IMAGE:figures/full_fig_p006_7.png]
Figure 8
Figure 8. Figure 8: One Side Wrinkle geometry VIII. CONCLUSION AND FUTURE WORK This paper presents a novel approach to real-time steering control for Vine robots through surface wrinkle fabrication. Our implementation successfully demonstrates real-time pla￾nar steering control using a du…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

14 extracted references · 13 canonical work pages

  1. [9]

    A soft robot that navigates its environment through growth,

    E. W. Hawkes, L. H. Blumenschein, J. D. Greer, and A. M. Okamura, “A soft robot that navigates its environment through growth,” Science Robotics, vol. 2, no. 8, p. eaan3028, 2017

  2. [11]

    Novel growing robot with in- flatable structure and heat-welding rotation mechanism,

    Y . Satake, A. Takanishi, and H. Ishii, “Novel growing robot with in- flatable structure and heat-welding rotation mechanism,” IEEE/ASME Transactions on Mechatronics, vol. 25, no. 4, pp. 1869–1877, 2020

  3. [1]

    Grow-hose-i: A hose type rescue robot passing smoothly through narrow rubble spaces,

    I. Kiryu, H. TSUKAGOSHI, and A. Kitagawa, “Grow-hose-i: A hose type rescue robot passing smoothly through narrow rubble spaces,” in Proceedings of the JFPS International Symposium on Fluid Power , vol. 2008, no. 7-3. The Japan Fluid Power System Society, 2008, pp. 815–820

  4. [2]

    Design, modeling, control, and application of everting vine robots,

    L. H. Blumenschein, M. M. Coad, D. A. Haggerty, A. M. Okamura, and E. W. Hawkes, “Design, modeling, control, and application of everting vine robots,” Frontiers in Robotics and AI , vol. 7, p. 548266, 2020

  5. [3]

    A growing soft robot with climbing plant–inspired adaptive behaviors for naviga- tion in unstructured environments,

    E. Del Dottore, A. Mondini, N. Rowe, and B. Mazzolai, “A growing soft robot with climbing plant–inspired adaptive behaviors for naviga- tion in unstructured environments,” Science Robotics , vol. 9, no. 86, p. eadi5908, 2024

  6. [4]

    Human interface for teleoperated object ma- nipulation with a soft growing robot,

    F. Stroppa, M. Luo, K. Yoshida, M. M. Coad, L. H. Blumenschein, and A. M. Okamura, “Human interface for teleoperated object ma- nipulation with a soft growing robot,” in 2020 IEEE International Conference on Robotics and Automation (ICRA) . IEEE, 2020, pp. 726–732

  7. [5]

    Development and Testing of a Vine Robot for Urban Search and Rescue in Confined Rubble Environments

    Z. Zhou, Y . Wang, E. W. Hawkes, and C. Li, “Development and testing of a vine robot for urban search and rescue in confined rubble environments,” arXiv preprint arXiv:2409.10000 , 2024

  8. [6]

    Vine robots,

    M. M. Coad, L. H. Blumenschein, S. Cutler, J. A. R. Zepeda, N. D. Naclerio, H. El-Hussieny, U. Mehmood, J.-H. Ryu, E. W. Hawkes, and A. M. Okamura, “Vine robots,” IEEE Robotics & Automation Magazine, vol. 27, no. 3, pp. 120–132, 2020

Show all 14 references
  1. [7]

    Material scrunching enables working channels in miniaturized vine-inspired robots,

    C. Girerd, A. Alvarez, E. W. Hawkes, and T. K. Morimoto, “Material scrunching enables working channels in miniaturized vine-inspired robots,” IEEE Transactions on Robotics , 2024

  2. [8]

    3d electromagnetic reconfiguration enabled by soft continuum robots,

    L. T. Gan, L. H. Blumenschein, Z. Huang, A. M. Okamura, E. W. Hawkes, and J. A. Fan, “3d electromagnetic reconfiguration enabled by soft continuum robots,” IEEE Robotics and Automation Letters , vol. 5, no. 2, pp. 1704–1711, 2020

  3. [10]

    Hybrid vine robot with internal steering-reeling mechanism enhances system-level capabilities,

    D. A. Haggerty, N. D. Naclerio, and E. W. Hawkes, “Hybrid vine robot with internal steering-reeling mechanism enhances system-level capabilities,” IEEE Robotics and Automation Letters , vol. 6, no. 3, pp. 5437–5444, 2021

  4. [12]

    A soft, steerable continuum robot that grows via tip extension,

    J. D. Greer, T. K. Morimoto, A. M. Okamura, and E. W. Hawkes, “A soft, steerable continuum robot that grows via tip extension,” Soft robotics, vol. 6, no. 1, pp. 95–108, 2019

  5. [13]

    Supplementary data for vine robot research,

    H. Liu, K. Smith, Z. Yang, and M. Yim, “Supplementary data for vine robot research,” 2025, available at: https://drive.google.com/drive/ folders/1PWctSVIzyFiMTjk 3H BA3H 5TGc tL6?usp=drive link

  6. [14]

    On curves of minimal length with a constraint on average curvature, and with prescribed initial and terminal positions and tangents,

    L. E. Dubins, “On curves of minimal length with a constraint on average curvature, and with prescribed initial and terminal positions and tangents,” American Journal of mathematics , vol. 79, no. 3, pp. 497–516, 1957

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.