Pith. sign in

REVIEW 1 major objections 31 references

Hybrid Data-Driven Predictive Control for Robust and Reactive Exoskeleton Locomotion Synthesis

T0 review · 1 major / 0 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read A single Hankel-matrix model can plan both foot contacts and body trajectory for an exoskeleton in real time, the paper claims.

desk verdict The submitted full text is an unrelated materials-science paper, so the exoskeleton controller cannot be evaluated on the evidence provided. read the letter →

arxiv 2508.10269 v1 pith:5STVGSMX submitted 2025-08-14 cs.RO

Kejun Li , Jeeseop Kim , Maxime Brunet , Marine Pétriaux , Yisong Yue , Aaron D. Ames This is my paper · ORCID
classification cs.RO
keywords hybriddata-drivenpredictivecontrolexoskeletonlocomotioncontactschedulingHankelmatrixstep-to-steptransitionsonlinereplanningbipedalwalking
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 hybrid data-driven predictive control (HDDPC), an extension of data-enabled predictive control that simultaneously plans foot-contact schedules and continuous-domain walking trajectories. It claims that a Hankel matrix built from step-to-step transition data can serve as the system model, enabling online replanning of both when and where to step as the environment changes. The intended value is an efficient, unified locomotion synthesis method for exoskeletons that reacts to disturbances without needing an explicit dynamics model. The provided full text, however, is an unrelated manuscript on crystal polymorphism and contains no derivation, data, or experimental validation for the exoskeleton claim; the abstract is the only statement of the contribution.

What carries the argument

The central object is the Hankel matrix-based representation of system dynamics, formed from recorded step-to-step (S2S) transitions. It acts as a non-parametric, data-driven model: a batch of past input-output trajectories is arranged in Hankel form, and the predictive controller uses that matrix to predict future behavior and to plan foot-contact schedules together with continuous trajectories. This is what lets the framework avoid an explicit analytical model while still considering discrete contact changes.

What would settle it

Run HDDPC on the Atalante exoskeleton over a disturbance sequence that changes required foot placement, then compare its predicted next state and contact timing with measured values. If the Hankel-model prediction error grows beyond a stability margin as the terrain changes, or if online replanning cannot meet the step cycle deadline, the central claim is falsified. Since the provided text contains no data, the decisive observation is the Atalante experiment itself and whether it demonstrates improved robustness over a baseline.

Watch

Extended reading notes

Core claim

On the basis of the abstract, the central claim is that robust and reactive exoskeleton walking can be produced by extending data-enabled predictive control to hybrid systems. HDDPC represents the dynamics with a Hankel matrix assembled from step-to-step transition data, and in the same optimization it chooses the foot contact sequence and the continuous-domain trajectories. This unifies contact scheduling with trajectory planning, so the exoskeleton can replan its gait online. The paper states this approach was validated on the Atalante exoskeleton, showing improved robustness and adaptability. The submitted full text does not contain that validation or the derivation; it is a different pap

Load-bearing premise

The framework works only if a finite batch of step-to-step transition data, arranged in a Hankel matrix, is a faithful enough model of the hybrid human-exoskeleton dynamics to choose safe new foot-contact schedules online.

Editorial extensions

If this is right

  • If the framework works as claimed, an exoskeleton can change its foot placement and step timing online in response to terrain or disturbances, rather than executing a fixed gait cycle.
  • Foot-contact scheduling and trajectory optimization become a single predictive-control problem, potentially removing the need to hand-design a separate high-level planner for step locations.
  • The approach could transfer to new users or new exoskeletons by collecting a fresh batch of step-to-step transition data, without deriving a new dynamics model for each system.
  • Because the model comes from data, the same framework may extend to other hybrid locomotion systems, such as bipedal robots or assistive devices, where contact transitions dominate the dynamics.

Reading between the lines

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

  • A testable extension would be to measure how HDDPC's prediction accuracy degrades when the exoskeleton encounters terrains or speeds not represented in the Hankel matrix; the claim of robustness depends on the data batch covering the relevant operating envelope.
  • The submitted text contains no experimental section, so the Atalante validation is currently an assertion rather than a demonstrated result; an independent check would require the actual trial data, disturbance protocol, and a comparison against a baseline controller.
  • If the Hankel matrix must be re-collected for each user or each exoskeleton, the practical advantage over model-based control will depend on how little data is needed and how quickly online replanning can run; those quantities are not yet specified in the provided text.
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

1 major / 0 minor

Summary. The manuscript under review, arXiv:2508.10269, is titled "Hybrid Data-Driven Predictive Control for Robust and Reactive Exoskeleton Locomotion Synthesis" and its abstract claims a new HDDPC framework that extends data-enabled predictive control, uses Hankel-matrix dynamics with step-to-step transitions, integrates contact scheduling with trajectory planning, and validates on the Atalante exoskeleton. However, the supplied full text is a completely different paper, arXiv:2508.10270, titled "Data-Driven Topological Analysis of Polymorphic Crystal Structures," by a different author group. That full text contains no derivation, algorithm, equations, or experimental results related to exoskeletons, predictive control, or HDDPC. As submitted, the only HDDPC-related material is the abstract itself; the body provides no supporting evidence for any of the abstract's claims.

Significance. If the claimed HDDPC framework were properly presented, the topic would be relevant to the robotics community: extending data-enabled predictive control to hybrid, contact-rich exoskeleton locomotion, with simultaneous contact-schedule and trajectory planning, could be a meaningful contribution. The framework's algorithmic novelty and its validation on a real exoskeleton would be its main strengths. However, the submitted artifact does not contain those contributions. The scientific significance cannot be assessed because no method, derivations, or experimental data are present. The manuscript in its current form is not a paper about the claimed topic; it is an unrelated materials-science manuscript, so the claimed contribution is unsupported.

major comments (1)
  1. [Entire manuscript] The manuscript body is unrelated to the claimed topic. The abstract describes HDDPC for exoskeleton locomotion, but the full text is a materials-science paper on polymorphic crystal structures, with no equations, algorithms, or experiments on predictive control, Hankel matrices, contact scheduling, or the Atalante exoskeleton. Consequently, the central claim that HDDPC extends data-enabled predictive control and is validated on an exoskeleton has no evidentiary basis in the submitted text. This is a load-bearing failure that cannot be remedied by minor revision.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity identifiable: the supplied full text does not contain the HDDPC derivation, algorithm, or Atalante experiments, so there is no derivation chain to reduce to its inputs.

full rationale

The claimed paper (arXiv:2508.10269) promises a hybrid data-driven predictive control framework for exoskeletons, but the accompanying full text is an unrelated condensed-matter manuscript on polymorphic crystal structures (arXiv:2508.10270). No equations, Hankel-matrix formulation, S2S transition model, contact-scheduling optimization, or experimental section for the Atalante exoskeleton are present. The only HDDPC-related content is the abstract itself. Circularity analysis requires an actual derivation chain whose predicted outputs can be compared against fitted inputs or self-citations. Here there is no method text, no fitted parameter, and no cited uniqueness theorem: the central claim is unverifiable from the supplied artifact, but unverifiability is not circularity. Under the hard rule that circularity may be flagged only when the specific reduction can be quoted, no circular step can be exhibited, so the appropriate score is 0. Concerns about the mismatch between abstract and body are correctness/completeness risks, not circularity.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

Scores and ledger are derived from the abstract alone because the full text is a different paper. The listed axioms are implicit modeling assumptions in the method's design, not explicit postulates. No free parameters or invented entities are named in the abstract.

assumptions (3)
  • domain assumption A Hankel matrix-based representation can model the hybrid dynamics of an exoskeleton-human system for online control.
    The abstract states the method 'utilizes a Hankel matrix-based representation to model system dynamics', but no justification or validation is available in the submitted text.
  • domain assumption Incorporating step-to-step transitions enhances adaptability in dynamic environments.
    The abstract asserts this improvement without supporting analysis or experiments in the provided text.
  • domain assumption The Atalante exoskeleton experiments are a valid demonstration of robustness and reactivity.
    The abstract claims validation on the Atalante exoskeleton, but no experimental details are present in the provided full text.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Hybrid Data-Driven Predictive Control for Robust and Reactive Exoskeleton Locomotion Synthesis." pith.science (2026). https://pith.science/paper/5STVGSMX

@misc{pith2026250810269,
  author       = {Pith},
  title        = {Pith review of: Hybrid Data-Driven Predictive Control for Robust and Reactive Exoskeleton Locomotion Synthesis},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5STVGSMX}},
  note         = {Machine review of arXiv:2508.10269}
}
read the original abstract

Robust bipedal locomotion in exoskeletons requires the ability to dynamically react to changes in the environment in real time. This paper introduces the hybrid data-driven predictive control (HDDPC) framework, an extension of the data-enabled predictive control, that addresses these challenges by simultaneously planning foot contact schedules and continuous domain trajectories. The proposed framework utilizes a Hankel matrix-based representation to model system dynamics, incorporating step-to-step (S2S) transitions to enhance adaptability in dynamic environments. By integrating contact scheduling with trajectory planning, the framework offers an efficient, unified solution for locomotion motion synthesis that enables robust and reactive walking through online replanning. We validate the approach on the Atalante exoskeleton, demonstrating improved robustness and adaptability.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

31 extracted references · 29 canonical work pages

  1. [2]

    Harib, A

    O. Harib, A. Hereid, A. Agrawal, T. Gurriet, S. Finet, G. Boeris, A. Duburcq, M. E. Mungai, M. Masselin, A. D. Ames et al., ``Feedback control of an exoskeleton for paraplegics: Toward robustly stable, hands-free dynamic walking,'' IEEE Control Systems Magazine, vol. 38, no. 6, pp. 61--87, 2018

    work page 2018

  2. [3]

    Gurriet, S

    T. Gurriet, S. Finet, G. Boeris, A. Duburcq, A. Hereid, O. Harib, M. Masselin, J. Grizzle, and A. D. Ames, ``Towards restoring locomotion for paraplegics: Realizing dynamically stable walking on exoskeletons,'' in 2018 IEEE international conference on robotics and automation (ICRA). 1em plus 0.5em minus 0.4em IEEE, 2018, pp. 2804--2811

    work page 2018

  3. [4]

    Kerdraon, J

    J. Kerdraon, J. G. Previnaire, M. Tucker, P. Coignard, W. Allegre, E. Knappen, and A. Ames, ``Evaluation of safety and performance of the self balancing walking system atalante in patients with complete motor spinal cord injury,'' Spinal cord series and cases, vol. 7, no. 1, p. 71, 2021

    work page 2021

  4. [5]

    J. W. Grizzle, C. Chevallereau, R. W. Sinnet, and A. D. Ames, ``Models, feedback control, and open problems of 3D bipedal robotic walking,'' Automatica, vol. 50, no. 8, pp. 1955--1988, 2014

    work page 1955

  5. [6]

    J. P. Reher, A. Hereid, S. Kolathaya, C. M. Hubicki, and A. D. Ames, ``Algorithmic foundations of realizing multi-contact locomotion on the humanoid robot durus,'' in Algorithmic Foundations of Robotics XII: Proceedings of the Twelfth Workshop on the Algorithmic Foundations of Robotics. 1em plus 0.5em minus 0.4em Springer, 2020, pp. 400--415

    work page 2020

  6. [7]

    E. R. Westervelt, J. W. Grizzle, and D. E. Koditschek, ``Hybrid zero dynamics of planar biped walkers,'' IEEE transactions on automatic control, vol. 48, no. 1, pp. 42--56, 2003

    work page 2003

  7. [8]

    Reher and A

    J. Reher and A. D. Ames, ``Dynamic walking: Toward agile and efficient bipedal robots,'' Annual Review of Control, Robotics, and Autonomous Systems, vol. 4, pp. 535--572, 2021

    work page 2021

  8. [9]

    Kajita, F

    S. Kajita, F. Kanehiro, K. Kaneko, K. Fujiwara, K. Harada, K. Yokoi, and H. Hirukawa, ``Biped walking pattern generation by using preview control of zero-moment point,'' in 2003 IEEE international conference on robotics and automation (Cat. No. 03CH37422), vol. 2. 1em plus 0.5em minus 0.4em IEEE, 2003, pp. 1620--1626

    work page 2003

Show all 31 references
  1. [10]

    P. M. Wensing and D. E. Orin, ``High-speed humanoid running through control with a 3d-slip model,'' in 2013 IEEE/RSJ International Conference on Intelligent Robots and Systems. 1em plus 0.5em minus 0.4em IEEE, 2013, pp. 5134--5140

    2013

  2. [11]

    Englsberger, C

    J. Englsberger, C. Ott, and A. Albu-Sch \"a ffer, ``Three-dimensional bipedal walking control based on divergent component of motion,'' Ieee transactions on robotics, vol. 31, no. 2, pp. 355--368, 2015

    2015

  3. [12]

    Hoeller, N

    D. Hoeller, N. Rudin, D. Sako, and M. Hutter, ``Anymal parkour: Learning agile navigation for quadrupedal robots,'' Science Robotics, vol. 9, no. 88, p. eadi7566, 2024

    2024

  4. [13]

    Z. Li, X. B. Peng, P. Abbeel, S. Levine, G. Berseth, and K. Sreenath, ``Reinforcement learning for versatile, dynamic, and robust bipedal locomotion control,'' arXiv preprint arXiv:2401.16889, 2024

    2024 arXiv

  5. [14]

    Markovsky, J

    I. Markovsky, J. C. Willems, S. Van Huffel, and B. De Moor, Exact and approximate modeling of linear systems: A behavioral approach. 1em plus 0.5em minus 0.4em SIAM, 2006

    2006

  6. [15]

    Coulson, J

    J. Coulson, J. Lygeros, and F. D \"o rfler, ``Data-enabled predictive control: In the shallows of the deepc,'' in 2019 18th European Control Conference (ECC). 1em plus 0.5em minus 0.4em IEEE, 2019, pp. 307--312

    2019

  7. [16]

    R. T. Fawcett, K. Afsari, A. D. Ames, and K. Akbari Hamed, ``Toward a data-driven template model for quadrupedal locomotion,'' IEEE Robotics and Automation Letters, vol. 7, no. 3, pp. 7636--7643, 2022

    2022

  8. [17]

    R. T. Fawcett, L. Amanzadeh, J. Kim, A. D. Ames, and K. Akbari Hamed, ``Distributed data-driven predictive control for multi-agent collaborative legged locomotion,'' in 2023 IEEE International Conference on Robotics and Automation (ICRA). 1em plus 0.5em minus 0.4em IEEE, 2023,...

    2023

  9. [18]

    Ulkir, G

    O. Ulkir, G. Akgun, A. Nasab, and E. Kaplanoglu, ``Data-driven predictive control of a pneumatic ankle foot orthosis.'' Advances in Electrical & Computer Engineering, vol. 21, no. 1, 2021

    2021

  10. [19]

    B. J. Stephens and C. G. Atkeson, ``Push recovery by stepping for humanoid robots with force controlled joints,'' in 2010 10th IEEE-RAS International conference on humanoid robots. 1em plus 0.5em minus 0.4em IEEE, 2010, pp. 52--59

    2010

  11. [20]

    Xiong and A

    X. Xiong and A. Ames, ``3-d underactuated bipedal walking via h-lip based gait synthesis and stepping stabilization,'' IEEE Transactions on Robotics, vol. 38, no. 4, pp. 2405--2425, 2022

    2022

  12. [21]

    A. W. Winkler, C. D. Bellicoso, M. Hutter, and J. Buchli, ``Gait and trajectory optimization for legged systems through phase-based end-effector parameterization,'' IEEE Robotics and Automation Letters, vol. 3, no. 3, pp. 1560--1567, 2018

    2018

  13. [22]

    Seyde, J

    T. Seyde, J. Carius, R. Grandia, F. Farshidian, and M. Hutter, ``Locomotion planning through a hybrid bayesian trajectory optimization,'' in 2019 International Conference on Robotics and Automation (ICRA). 1em plus 0.5em minus 0.4em IEEE, 2019, pp. 5544--5550

    2019

  14. [23]

    K. Li, J. Kim, X. Xiong, K. A. Hamed, Y. Yue, and A. D. Ames, ``Data-driven predictive control for robust exoskeleton locomotion,'' in 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). 1em plus 0.5em minus 0.4em IEEE, 2024, pp. 162--169

    2024

  15. [24]

    Morris and J

    B. Morris and J. W. Grizzle, ``A restricted poincar \'e map for determining exponentially stable periodic orbits in systems with impulse effects: Application to bipedal robots,'' in Proceedings of the 44th IEEE Conference on Decision and Control. 1em plus 0.5em minus 0.4em IEE...

    2005

  16. [25]

    Kajita, F

    S. Kajita, F. Kanehiro, K. Kaneko, K. Fujiwara, K. Yokoi, and H. Hirukawa, ``A realtime pattern generator for biped walking,'' in Proceedings 2002 IEEE International Conference on Robotics and Automation (Cat. No. 02CH37292), vol. 1. 1em plus 0.5em minus 0.4em IEEE, 2002, pp. 31--37

    2002

  17. [26]

    H. J. Van Waarde, C. De Persis, M. K. Camlibel, and P. Tesi, ``Willems’ fundamental lemma for state-space systems and its extension to multiple datasets,'' IEEE Control Systems Letters, vol. 4, no. 3, pp. 602--607, 2020

    2020

  18. [27]

    o hler, M. A. M \

    J. Berberich, J. K \"o hler, M. A. M \"u ller, and F. Allg \"o wer, ``Data-driven model predictive control with stability and robustness guarantees,'' IEEE Transactions on Automatic Control, vol. 66, no. 4, pp. 1702--1717, 2020

    2020

  19. [28]

    J. C. Willems, P. Rapisarda, I. Markovsky, and B. L. De Moor, ``A note on persistency of excitation,'' Systems & Control Letters, vol. 54, no. 4, pp. 325--329, 2005

    2005

  20. [29]

    W \"a chter and L

    A. W \"a chter and L. T. Biegler, ``On the implementation of an interior-point filter line-search algorithm for large-scale nonlinear programming,'' Mathematical programming, vol. 106, pp. 25--57, 2006

    2006

  21. [30]

    Stellato, G

    B. Stellato, G. Banjac, P. Goulart, A. Bemporad, and S. Boyd, `` OSQP : an operator splitting solver for quadratic programs,'' Mathematical Programming Computation, vol. 12, no. 4, pp. 637--672, 2020

    2020

  22. [31]

    Todorov, T

    E. Todorov, T. Erez, and Y. Tassa, ``Mujoco: A physics engine for model-based control,'' in 2012 IEEE/RSJ international conference on intelligent robots and systems. 1em plus 0.5em minus 0.4em IEEE, 2012, pp. 5026--5033

    2012

  23. [32]

    Supplementary video: https://www.youtube.com/watch?v=OmVjbQzUGQ0,

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.