Pith. sign in

REVIEW 4 major objections 8 minor 2 cited by

Fast and Modular Whole-Body Lagrangian Dynamics of Legged Robots with Changing Morphology

T0 review · 4 major / 8 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read By embedding binary link and leg existence numbers in the mass and Coriolis matrices, the equations of motion of a damaged legged robot are obtained by matrix addition, subtraction, and row/column removal—no re-derivation or retraining.

desk verdict The modular mass/Coriolis decomposition is a real idea, but the gravity vector breaks the row-deletion story, so the central "any damage" claim is not supported as written. read the letter →

arxiv 2504.16383 v1 pith:THXBXTNZ submitted 2025-04-23 cs.RO nlin.AO

Sahand Farghdani , Omar Abdelrahman , Robin Chhabra This is my paper · ORCID
classification cs.ROnlin.AO MSC 70E6070E5593C85
keywords multi-leggedrobotswhole-bodydynamicsBoltzmann-Hamelequationsscrewtheorymodularmodelingmorphologychangedamageadaptationreal-timesimulation
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 seeks to establish that whole-body dynamics of legged robots can be written so that damage to a leg or link is handled by flipping binary switches in the equations of motion rather than rebuilding the model. The authors derive a modular form of the Boltzmann-Hamel equations in which link existence numbers and leg existence numbers are embedded in the mass and Coriolis matrices. Removing a leg is then a matrix operation: delete the corresponding blocks and rows and columns, and the reduced-order equations follow without differentiation. If the claim is right, a robot that loses limbs in the field can update the model its controller uses during the same control cycle, with no retraining and no manual recalculation. The paper supports the claim with a hexapod simulation that matches inertial and camera measurements and runs about three times faster than real time.

What carries the argument

The central object is the modular mass matrix (27) together with the modular Coriolis matrix (36), built from the Boltzmann-Hamel equations (singularity-free Lagrange equations on the Lie group SE(3)) and screw-theory twists. The mechanism that carries the argument is block decoupling: under the single-branch serial-leg assumption, each leg's Jacobian has nonzero columns only for that leg's own joints (Lemma 2), so every link's contribution to the mass matrix is nonzero only in that leg's block. Binary link and leg existence numbers $x_{ij}$ and $\bar{x}_i$ multiply these blocks, so a damaged morphology is exactly the healthy matrix with some blocks zeroed and then rows and columns deleted. The Coriolis derivative terms are precomputed symbolically by decomposing each mass block into constant coefficient matrices times a fixed vector of trigonometric functions (46), so runtime updates require only substitution and matrix operations.

What would settle it

Take a robot with one branched or closed-chain leg and run the same construction: assemble the modular mass matrix (27) from per-link blocks, then compare it term by term with a direct re-derivation of the full equations for that leg. If off-diagonal coupling between joints in different branches is missing, the modularity claim is confined to unbranched serial legs; if the matrices match and the damaged-model update still runs faster than a full re-derivation, the central claim survives this test.

Watch

Extended reading notes

Core claim

The paper's central claim is that the whole-body dynamics of a multi-legged robot can be made modular in morphology: the presence of each link and each leg is coded as a binary existence number ($x_{ij}$ and $\bar{x}_i$) inside the block-sparse mass matrix (27) and Coriolis matrix (36). When a link or leg is lost, the equations of motion for the damaged robot are obtained from the healthy model by matrix addition, subtraction, and deleting rows and columns; no symbolic differentiation, re-derivation, or retraining is needed. The authors also claim that the same modular equations run about three times faster than real time on a standard CPU, and they validate the model against motion-capture and inertial measurements from a hexapod under two different leg-damage patterns.

Load-bearing premise

The whole modular assembly depends on each leg being a simple unbranched chain of joints connected to one rigid body, so that a leg's motion depends only on that leg's own joints; if a leg branches, forms a loop, or the body bends, the block-switch construction no longer represents the robot.

Editorial extensions

If this is right

  • A control loop can reconfigure its dynamics model mid-simulation as soon as a leg or link failure is detected, without pausing for re-derivation.
  • Model-based damage identification becomes a comparison among candidate morphology vectors $\bar{X}$ and $X$: each candidate's equations are assembled by the same matrix operations and scored against measured motion.
  • Because the equations remain singularity-free on SE(3), the model stays valid during extreme body poses that damage can cause, where Euler-angle parameterizations fail.
  • New leg morphologies need to be added to the symbolic repository only once; any number of instances attached at arbitrary body locations reuse the same precomputed matrices.
  • The same symbolic model generation can support online planning over multiple candidate structures without a separate training phase.

Reading between the lines

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

  • Beyond the paper: the existence-number switch could be used in reverse, treating $\bar{X}$ and $X$ as hidden binary states and inferring the most probable morphology from residuals between predicted and measured accelerations, giving online self-modeling without retraining.
  • Beyond the paper: the same block-decoupling property should extend to adding legs, not just removing them, as long as the new leg's morphology is already in the repository; that would cover self-reconfiguration and modular assembly, not only damage.
  • Beyond the paper: a testable extension is to stress the model in regimes the paper does not cover, such as asymmetric ground contacts, slipping feet, or high-speed gaits where leg inertia and Coriolis coupling matter more; the runtime advantage may shrink as contact nonlinearities dominate.
  • Beyond the paper: the method's dependence on single-branch serial legs suggests a natural benchmark—quantify the error if a branched or closed-chain leg is approximated by this assembly, which would identify the boundary of the modularity result.
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 / 8 minor

Summary. The paper proposes a modular whole-body dynamic modeling framework for multi-legged robots (MLRs) based on Boltzmann-Hamel equations and screw theory. The authors derive modular forms of the mass matrix (Proposition 1) and Coriolis matrix (Proposition 2) that embed link and leg existence numbers, and a fast symbolic decomposition that reuses per-leg expressions for arbitrary leg morphologies. They claim that morphology changes caused by damage can be accommodated through matrix addition, subtraction, and row/column removal without re-deriving the equations of motion. Validation is provided via a custom simulation engine and hardware experiments on a Hiwonder JetHexa hexapod under leg and link damage scenarios, with a reported computation speed about three times faster than real-time.

Significance. If the central claim were fully supported, the framework would be a valuable contribution to real-time model-based control and damage adaptation for legged robots, as it would avoid re-deriving the equations of motion or retraining learned models after structural damage. The algebraic derivation of the modular mass and Coriolis matrices is original and appears internally consistent under the stated single-branch serial-chain assumption; the closed-form derivative expressions in Lemma 3 are a useful standalone contribution. The use of hardware experiments, albeit qualitative, is a strength, and the symbolic decomposition approach is a reasonable basis for a fast implementation. However, the potential-force modularity gap and the unaddressed non-suffix link-removal cases are load-bearing limitations that currently prevent the paper from supporting its advertised 'any combination' claim.

major comments (4)
  1. [Section IV.C, Algorithms 1 and 3] The potential-force vector N(Q) in Eq. (9) is never given a modular form. Algorithms 1 (lines 25-28) and 3 (lines 29-32) only delete the rows of N that correspond to removed links or legs. This is not equivalent to recomputing N for the damaged morphology: the body row N_b and the rows of retained joints receive gravitational contributions from every distal link through ∂U/∂θ_ik, so a removed link's contribution remains in those rows. For example, in the second damage scenario where l42 and l43 are removed, the retained θ41 row of the original N still contains the gravitational torques from links 2 and 3 unless U is recomputed. The same applies to the body row when entire legs are removed. Consequently, the reduced equations of motion implemented by the algorithms are not the dynamics of the damaged robot, and the central claim that 'any combination' of missing links can be handled by row/column removal is not supported.
  2. [Lemma 2, Eq. (16) and Definition 1] The modular mass matrix of Proposition 1 is only correct for suffix removals of links. By Eq. (16), the Jacobian of a distal link l_ij has nonzero columns only for joints 1..j, and the instantaneous twists ξ'_iβ for β>k are functions of the removed coordinate θ_ik through Eq. (17). If a non-suffix link is removed while a distal link remains, the retained entries of M_ij for that distal link still contain the removed coordinate after row/column deletion, so the reduced mass matrix is not a valid model of the damaged leg. The manuscript never defines what 'a valid morphology' means in Algorithm 1 line 5; if only suffix removals are allowed, that restriction should be stated and the 'any combination' wording in the abstract and introduction should be narrowed.
  3. [Section VI.B and Figures 9-10] The experimental section explicitly states that two damage scenarios are simulated, yet Figures 9 and 10 show results for a 'third damage scenario' that is never described in the text. The missing scenario makes the validation section incomplete and confusing; the authors should either add the corresponding description and setup or remove the orphaned figures.
  4. [Section VI.B, Table III] The validation of the model against hardware is purely qualitative: the paper presents time plots of body orientation and CoM position but no error metrics such as RMSE, maximum deviation, or correlation. Because the contact parameters in Table III are experimentally identified on the same robot, the agreement could be substantially influenced by parameter fitting. To support the claimed 'accuracy and adaptability', the paper should include quantitative error measures and ideally a sensitivity analysis with respect to the fitted contact parameters.
minor comments (8)
  1. [Section III.A] The text introduces the 'Special Euclidean grop'; this should be 'group'.
  2. [Eq. (12)] Eq. (12) says 'PEO formula'; the standard name is 'POE' (product of exponentials).
  3. [Reference [73]] Reference [73] appears to have a formatting error ('J. M. Scheurle and J'); please correct the author list.
  4. [Algorithm 1 line 5 and Algorithm 3 line 5] The phrase 'Xlegi is a valid morphology' is used without defining what makes a morphology valid; please specify the condition (e.g., suffix-preserving or arbitrary).
  5. [Section V, Eq. (46)] The decomposition in Eq. (46) assumes linear independence of the elements of Fα; the manuscript should justify this assumption for the considered leg morphologies or note any dependence.
  6. [Section VI.B] The reported runtime (5 s simulation in 1.93 s) is a single measurement without statistical spread or a baseline comparison; please state the number of runs and specify the comparison method.
  7. [Section I.A] The introduction defines damage to include 'locked joints', but the modular derivations only address link/leg removal; clarify how locked joints are represented in the model.
  8. [Figures 3-10] Figures 3-10 would benefit from axis labels and units; several plots appear compressed and hard to read.

Circularity Check

1 steps flagged · score 4.0 of 10

Foundational EOMs are imported from the authors' own prior work without proof, but the modular rearrangement itself is independently derived.

  1. self citation load bearing [Section III.A, Theorem 1 (Eqs. 6-11); used in Section IV, Propositions 1 and 2]
    "Based on this mathematical notation, in our recent publication [92], we presented a theory for deriving singularity-free whole-body equations for an MLR that will be presented in the next section. ... Theorem 1 (Boltzmann-Hamel equations of legged robots [92]). Consider an MLR ... The singularity-free Boltzmann-Hamel equations of the system are given by: M(θ) ˙v +C(θ,v)v +N (Q) =τ +JTFt"

    The modular mass matrix (Prop. 1, Eq. 27) and modular Coriolis matrix (Prop. 2, Eq. 36) are generated by inserting link/leg existence numbers into the block sums of Theorem 1's Eqs. (7) and (10). Theorem 1 itself is stated without proof and attributed to the authors' own prior publication [92]; no independent derivation, machine-checking, or reproduction is provided in this manuscript. Thus the first-principles content of the entire modular chain is inherited from an unverified self-citation. The modularity reconstruction itself is new and rests on Lemmas 1-3, so the circularity is partial rather than total.

full rationale

The paper's central modularity claim has substantial independent mathematical content: Lemmas 1-3, Proposition 1, and Proposition 2 work directly from the structure of the Jacobian blocks and do not merely rename the input. The main circularity concern is that the governing equations (Theorem 1) are taken from the authors' own prior paper [92] without proof in this text, making the foundation a load-bearing self-citation. The hardware validation uses contact parameters identified on the same robot (Table III), which weakens the validation as an external check, although it is not a fitted parameter renamed as a prediction. Separately, the potential-force vector N is handled only by row deletion in Algorithms 1 and 3 rather than by a modular recomputation of ∂U/∂θ; this is a correctness/completeness gap in the 'any combination of missing links' claim, not a circularity. External citations such as [67] and [93] are standard or external and are not treated as circular. On balance, the modular derivation is self-contained enough for a moderate score: some self-citation, but the central modularity claim retains independent content.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The central claim rests on the authors' prior theorem, the serial-branch leg structure, a known trigonometric structure result, a constant gravity assumption, and a linear spring-damper contact model with fitted parameters. The morphology vectors themselves are algorithmic, not physical.

free parameters (3)
  • Ground stiffness k_z = 10000 N/m
    Contact normal force spring constant, listed in Table III as experimentally identified; used in the contact model that produces the simulated motion compared to hardware.
  • Ground damping d_z = 150 N/(m/s)
    Contact normal damping coefficient, Table III, experimentally identified.
  • Viscous friction coefficient d_t = 50 N/(m/s)
    Tangential friction parameter, Table III, experimentally identified.
assumptions (5)
  • domain assumption Theorem 1 of [92]: the Boltzmann-Hamel equations (6)-(11) correctly describe the whole-body dynamics of an MLR with a 6-DoF main body and serial legs.
    Stated in Section III without proof, citing the authors' own prior work [92]; the entire modular derivation in Sections IV-V builds on these equations.
  • domain assumption Each leg is a single-branch serial chain attached to one rigid main body (Remark 1), so each leg's Jacobian has nonzero columns only for that leg's joints (Lemma 2).
    Introduced in Section IV Remark 1 and used throughout to establish the block-sparse modular mass matrix (27). This excludes parallel/closed-chain legs or flexible bodies; the authors call it WLOG but it is a real restriction.
  • standard math The leg mass matrices contain only trigonometric terms up to order 2n, as in Lemma 4 from [93].
    Used in Section V to precompute the I and T decomposition matrices; a known property of serial manipulators with revolute joints.
  • domain assumption Constant gravitational field (Remark 8).
    Required so that the gravity vector N can be precomputed symbolically per leg morphology and then evaluated at runtime; fails for non-uniform fields.
  • domain assumption Contact forces at leg tips follow a linear spring-damper normal model and viscous tangential friction.
    Used in Section VI to generate the simulated motion compared against hardware; the parameters are fitted to the hardware, so the validation is conditional on this model.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Fast and Modular Whole-Body Lagrangian Dynamics of Legged Robots with Changing Morphology." pith.science (2026). https://pith.science/paper/THXBXTNZ

@misc{pith2026250416383,
  author       = {Pith},
  title        = {Pith review of: Fast and Modular Whole-Body Lagrangian Dynamics of Legged Robots with Changing Morphology},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/THXBXTNZ}},
  note         = {Machine review of arXiv:2504.16383}
}
read the original abstract

Fast and modular modeling of multi-legged robots (MLRs) is essential for resilient control, particularly under significant morphological changes caused by mechanical damage. Conventional fixed-structure models, often developed with simplifying assumptions for nominal gaits, lack the flexibility to adapt to such scenarios. To address this, we propose a fast modular whole-body modeling framework using Boltzmann-Hamel equations and screw theory, in which each leg's dynamics is modeled independently and assembled based on the current robot morphology. This singularity-free, closed-form formulation enables efficient design of model-based controllers and damage identification algorithms. Its modularity allows autonomous adaptation to various damage configurations without manual re-derivation or retraining of neural networks. We validate the proposed framework using a custom simulation engine that integrates contact dynamics, a gait generator, and local leg control. Comparative simulations against hardware tests on a hexapod robot with multiple leg damage confirm the model's accuracy and adaptability. Additionally, runtime analyses reveal that the proposed model is approximately three times faster than real-time, making it suitable for real-time applications in damage identification and recovery.

Figures

Figures reproduced from arXiv: 2504.16383 by the authors.

Figure 1
Figure 1. Frames assignment for the main body and Leg [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Hiwonder JetHexa robot with motion-tracking LEDs [PITH_FULL_IMAGE:figures/full_fig_p011_2.png] view at source ↗
Figure 4
Figure 4. Main body COM position during the first damage scenario [PITH_FULL_IMAGE:figures/full_fig_p012_4.png] view at source ↗
Figures from the paper (7 more)
Figure 3
Figure 3. Figure 3: Main body orientation during the first damage scenario [PITH_FULL_IMAGE:figures/full_fig_p012_3.png]
Figure 5
Figure 5. Figure 5: Snapshots of 1.625 seconds of model and robot motions for the first damage scenario [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]
Figure 6
Figure 6. Figure 6: Main body orientation during the second damage scenario [PITH_FULL_IMAGE:figures/full_fig_p013_6.png]
Figure 7
Figure 7. Figure 7: Main body COM position during the second damage scenario [PITH_FULL_IMAGE:figures/full_fig_p013_7.png]
Figure 8
Figure 8. Figure 8: Snapshots of 1.625 seconds of model and robot motions motion for the second damage scenario [PITH_FULL_IMAGE:figures/full_fig_p014_8.png]
Figure 9
Figure 9. Figure 9: Main body orientation during the third damage scenario [PITH_FULL_IMAGE:figures/full_fig_p014_9.png]
Figure 10
Figure 10. Figure 10: Main body COM position during the third damage scenario [PITH_FULL_IMAGE:figures/full_fig_p014_10.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Robust Embodied Self-Identification of Morphology in Damaged Multi-Legged Robots

    cs.RO 2025-06 conditional novelty 4.0 of 10

    A hexapod robot can identify its damaged legs by matching its IMU-measured body orientation against a fast simulator, using a genetic algorithm with an FFT-based filter, achieving 89% scenario-level accuracy in under ...

  2. Evolutionary Gait Reconfiguration in Damaged Legged Robots

    cs.RO 2025-06 conditional novelty 4.0 of 10

    A training-free pipeline that selects a stable gait and optimizes leg tip trajectories with differential evolution restored forward walking in a damaged hexapod.

Reference graph

Works this paper leans on

94 extracted references · 64 canonical work pages · cited by 2 Pith papers

  1. [1]

    Bigdog, the rough-terrain quadruped robot,

    M. Raibert, K. Blankespoor, G. Nelson, and R. Playter, “Bigdog, the rough-terrain quadruped robot,” IFAC Proceedings Volumes , vol. 41, no. 2, pp. 10 822–10 825, 2008

    2008

  2. [2]

    Spacebok: A dynamic legged robot for space exploration,

    P. Arm, R. Zenkl, P. Barton, L. Beglinger, A. Dietsche, L. Ferrazzini, E. Hampp, J. Hinder, C. Huber, D. Schaufelberger et al., “Spacebok: A dynamic legged robot for space exploration,” in International Confer- ence on Robotics and Automation (ICRA) , 2019, pp. 6288–6294

    2019

  3. [3]

    Scientific exploration of chal- lenging planetary analog environments with a team of legged robots,

    P. Arm, G. Waibel, J. Preisig, T. Tuna, R. Zhou, V . Bickel, G. Ligeza, T. Miki, F. Kehl, H. Kolvenbach et al., “Scientific exploration of chal- lenging planetary analog environments with a team of legged robots,” Science robotics, vol. 8, no. 80, p. eade9548, 2023

    2023

  4. [4]

    Spot® Boston Dynamics,

    Boston Dynamics, “Spot® Boston Dynamics,” https://www. bostondynamics.com/spot, 2015, [Accessed: November, 2022]

    2015

  5. [5]

    ANYmal X by ANYbotics,

    ANYbotics, “ANYmal X by ANYbotics,” https://www.anybotics.com/ anymal-ex-proof-inspection-robot/, 2022, [Accessed: November, 2022]

    2022

  6. [6]

    Legged robots for object manipulation: A review,

    Y . Gong, G. Sun, A. Nair, A. Bidwai, R. CS, J. Grezmak, G. Sartoretti, and K. A. Daltorio, “Legged robots for object manipulation: A review,” Frontiers in Mechanical Engineering , vol. 9, p. 1142421, 2023

    2023

  7. [7]

    Model predictive control of legged and humanoid robots: models and algorithms,

    S. Katayama, M. Murooka, and Y . Tazaki, “Model predictive control of legged and humanoid robots: models and algorithms,” Advanced Robotics, vol. 37, no. 5, pp. 298–315, 2023

    2023

  8. [8]

    Design and control of a miniature bipedal robot with proprioceptive actuation for dynamic behaviors,

    Y . Liu, J. Shen, J. Zhang, X. Zhang, T. Zhu, and D. Hong, “Design and control of a miniature bipedal robot with proprioceptive actuation for dynamic behaviors,” in 2022 International Conference on Robotics and Automation (ICRA). IEEE, 2022, pp. 8547–8553

    2022

Show all 94 references
  1. [9]

    Multicontact locomotion of legged robots,

    J. Carpentier and N. Mansard, “Multicontact locomotion of legged robots,” IEEE Transactions on Robotics , vol. 34, no. 6, pp. 1441–1460, 2018

    2018

  2. [10]

    Convex model predictive control of single rigid body model on so(3) for versatile dynamic legged motions,

    J. Shen and D. Hong, “Convex model predictive control of single rigid body model on so(3) for versatile dynamic legged motions,” in 2022 International Conference on Robotics and Automation (ICRA) , 2022, pp. 6586–6592

    2022

  3. [11]

    Optimizing bipedal maneuvers of single rigid-body models for rein- forcement learning,

    R. Batke, F. Yu, J. Dao, J. Hurst, R. L. Hatton, A. Fern, and K. Green, “Optimizing bipedal maneuvers of single rigid-body models for rein- forcement learning,” in 2022 IEEE-RAS 21st International Conference on Humanoid Robots (Humanoids) . IEEE, 2022, pp. 714–721

    2022

  4. [12]

    An investigation of 2 nd-order fixed point slip behavior,

    I. Kontolatis and E. Papadopoulos, “An investigation of 2 nd-order fixed point slip behavior,” in 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) . IEEE, 2018, pp. 1739–1744

    2018

  5. [13]

    Dynamic period-two gait generation in a hexapod robot based on the fixed-point motion of a reduced-order model,

    W.-C. Lu and P.-C. Lin, “Dynamic period-two gait generation in a hexapod robot based on the fixed-point motion of a reduced-order model,” in 2019 International Conference on Robotics and Automation (ICRA). IEEE, 2019, pp. 7427–7433

    2019

  6. [14]

    Dynamic modeling and gait analysis for miniature robots in the absence of foot placement control,

    M. Askari and O. ¨Ozcan, “Dynamic modeling and gait analysis for miniature robots in the absence of foot placement control,” in 2019 international conference on robotics and automation (ICRA) . IEEE, 2019, pp. 9754–9760

    2019

  7. [15]

    A framework for dynamic modeling of legged modular miniature robots with soft backbones,

    N. Mahkam and O. ¨Ozcan, “A framework for dynamic modeling of legged modular miniature robots with soft backbones,” Robotics and Autonomous Systems, vol. 144, p. 103841, 2021

    2021

  8. [16]

    Optimal feet-forces’ and torque distributions of six-legged robot maneuvering on various terrains,

    A. Mahapatra, S. S. Roy, and D. K. Pratihar, “Optimal feet-forces’ and torque distributions of six-legged robot maneuvering on various terrains,” Robotica, vol. 38, no. 6, p. 1041–1063, 2020

    2020

  9. [17]

    Analysis of position and angular velocity of four-legged robot (mini-bot) from dynamic model using euler-lagrange method,

    S. Y . Naing and T. Rain, “Analysis of position and angular velocity of four-legged robot (mini-bot) from dynamic model using euler-lagrange method,” in 2019 International Conference on Industrial Engineering, Applications and Manufacturing (ICIEAM) , 2019, pp. 1–4

    2019

  10. [18]

    Modeling of a quadruped robot with spine joints and full-dynamics simulation environment con- struction,

    O. K. Adak, B. Bahceci, and K. Erbatur, “Modeling of a quadruped robot with spine joints and full-dynamics simulation environment con- struction,” arXiv preprint arXiv:2203.09622 , 2022

    2022 arXiv

  11. [19]

    Optimization-based control for dynamic legged robots,

    P. M. Wensing, M. Posa, Y . Hu, A. Escande, N. Mansard, and A. Del Prete, “Optimization-based control for dynamic legged robots,” IEEE Transactions on Robotics , 2023

    2023

  12. [20]

    Biswal and P

    P. Biswal and P. K. Mohanty, “Modeling and effective foot force 14 Fig. 8. Snapshots of 1.625 seconds of model and robot motions motion for the second damage scenario Fig. 9. Main body orientation during the third damage scenario Fig. 10. Main body COM position during the thir...

  13. [21]

    Artificial neural network-based ground reaction force estimation and learning for dynamic-legged robot systems,

    J. An and I. Lee, “Artificial neural network-based ground reaction force estimation and learning for dynamic-legged robot systems,” PeerJ Computer Science, vol. 9, p. e1720, 2023

    2023

  14. [22]

    Fully spiking neural network for legged robots,

    X. Jiang, Q. Zhang, J. Sun, and R. Xu, “Fully spiking neural network for legged robots,” arXiv preprint arXiv:2310.05022 , 2023

    2023 arXiv

  15. [23]

    Data efficient reinforcement learning for legged robots,

    Y . Yang, K. Caluwaerts, A. Iscen, T. Zhang, J. Tan, and V . Sindhwani, “Data efficient reinforcement learning for legged robots,” in Conference on Robot Learning . PMLR, 2020, pp. 1–10

    2020

  16. [24]

    Energy- based legged robots terrain traversability modeling via deep inverse reinforcement learning,

    L. Gan, J. W. Grizzle, R. M. Eustice, and M. Ghaffari, “Energy- based legged robots terrain traversability modeling via deep inverse reinforcement learning,” IEEE Robotics and Automation Letters , vol. 7, no. 4, pp. 8807–8814, 2022

    2022

  17. [25]

    Learning agile and dynamic motor skills for legged robots,

    J. Hwangbo, J. Lee, A. Dosovitskiy, D. Bellicoso, V . Tsounis, V . Koltun, and M. Hutter, “Learning agile and dynamic motor skills for legged robots,” Science Robotics, vol. 4, no. 26, p. eaau5872, 2019

    2019

  18. [26]

    Generic neural locomotion control framework for legged robots,

    M. Thor, T. Kulvicius, and P. Manoonpong, “Generic neural locomotion control framework for legged robots,” IEEE transactions on neural networks and learning systems , vol. 32, no. 9, pp. 4013–4025, 2020

    2020

  19. [27]

    Legged robots that keep on learning: Fine-tuning locomotion policies in the real world,

    L. Smith, J. C. Kew, X. B. Peng, S. Ha, J. Tan, and S. Levine, “Legged robots that keep on learning: Fine-tuning locomotion policies in the real world,” in 2022 International Conference on Robotics and Automation (ICRA). IEEE, 2022, pp. 1593–1599

    2022

  20. [28]

    Legged robot running using a physics-data hybrid motion template,

    W.-S. Yang, W.-C. Lu, and P.-C. Lin, “Legged robot running using a physics-data hybrid motion template,” IEEE Transactions on Robotics , vol. 37, no. 5, pp. 1680–1695, 2021

    2021

  21. [29]

    Dynamic modeling, stability, and energy efficiency of a quadrupedal walking machine,

    B. S. Lin and S.-M. Song, “Dynamic modeling, stability, and energy efficiency of a quadrupedal walking machine,” Journal of Robotic Systems, vol. 18, no. 11, pp. 657–670, 2001

    2001

  22. [30]

    Dynamic modelling of a four-legged robot,

    M. Bennani and F. Giri, “Dynamic modelling of a four-legged robot,” Journal of Intelligent and Robotic Systems , vol. 17, pp. 419–428, 1996

    1996

  23. [31]

    Estimation of optimal feet forces and joint torques for on-line control of six-legged robot,

    S. S. Roy, A. K. Singh, and D. K. Pratihar, “Estimation of optimal feet forces and joint torques for on-line control of six-legged robot,” Robotics and Computer-Integrated Manufacturing , vol. 27, no. 5, pp. 910–917, 2011

    2011

  24. [32]

    Dynamic modeling, stability and energy consumption analysis of a realistic six-legged walking robot,

    S. S. Roy and D. K. Pratihar, “Dynamic modeling, stability and energy consumption analysis of a realistic six-legged walking robot,” Robotics and Computer-Integrated Manufacturing , vol. 29, no. 2, pp. 400–416, 2013

    2013

  25. [33]

    High speed trot-running: Im- plementation of a hierarchical controller using proprioceptive impedance control on the mit cheetah,

    D. J. Hyun, S. Seok, J. Lee, and S. Kim, “High speed trot-running: Im- plementation of a hierarchical controller using proprioceptive impedance control on the mit cheetah,” The International Journal of Robotics Research, vol. 33, no. 11, pp. 1417–1445, 2014

    2014

  26. [34]

    Centroidal dynamics of a humanoid robot,

    D. E. Orin, A. Goswami, and S.-H. Lee, “Centroidal dynamics of a humanoid robot,” Autonomous robots, vol. 35, pp. 161–176, 2013

    2013

  27. [35]

    Simultaneous contact, gait, and motion planning for robust multilegged locomotion via mixed-integer convex optimization,

    B. Aceituno-Cabezas, C. Mastalli, H. Dai, M. Focchi, A. Radulescu, D. G. Caldwell, J. Cappelletto, J. C. Grieco, G. Fern ´andez-L´opez, and C. Semini, “Simultaneous contact, gait, and motion planning for robust multilegged locomotion via mixed-integer convex optimization,” IEE...

    2017

  28. [36]

    Gait and trajectory optimization for legged systems through phase-based end-effector parameterization,

    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

  29. [37]

    Study of dynamic biped locomotion on rugged terrain-derivation and application of the linear inverted pendulum mode,

    S. Kajita and K. Tani, “Study of dynamic biped locomotion on rugged terrain-derivation and application of the linear inverted pendulum mode,” in Proceedings. 1991 IEEE International Conference on Robotics and Automation. IEEE Computer Society, 1991, pp. 1405–1406. 15

    1991

  30. [38]

    Online walking motion generation with automatic footstep placement,

    A. Herdt, H. Diedam, P.-B. Wieber, D. Dimitrov, K. Mombaur, and M. Diehl, “Online walking motion generation with automatic footstep placement,” Advanced Robotics, vol. 24, no. 5-6, pp. 719–737, 2010

    2010

  31. [39]

    Zmp support areas for multicontact mobility under frictional constraints,

    S. Caron, Q.-C. Pham, and Y . Nakamura, “Zmp support areas for multicontact mobility under frictional constraints,” IEEE Transactions on Robotics, vol. 33, no. 1, pp. 67–80, 2016

    2016

  32. [40]

    Trajectory planning of flexible walking for biped robots using linear inverted pendulum model and linear pendulum model,

    L. Li, Z. Xie, X. Luo, and J. Li, “Trajectory planning of flexible walking for biped robots using linear inverted pendulum model and linear pendulum model,” Sensors, vol. 21, no. 4, p. 1082, 2021

    2021

  33. [41]

    Drs-lip: Linear inverted pendulum model for legged locomotion on dynamic rigid surfaces,

    A. Iqbal, S. Veer, and Y . Gu, “Drs-lip: Linear inverted pendulum model for legged locomotion on dynamic rigid surfaces,” arXiv preprint arXiv:2202.00151, 2022

    2022 arXiv

  34. [42]

    Integrat- ing physics-based modeling with machine learning: A survey,

    J. Willard, X. Jia, S. Xu, M. Steinbach, and V . Kumar, “Integrat- ing physics-based modeling with machine learning: A survey,” arXiv preprint arXiv:2003.04919, vol. 1, no. 1, pp. 1–34, 2020

    2003 arXiv

  35. [43]

    Physics-guided deep learning for dynamical systems: A survey,

    R. Wang and R. Yu, “Physics-guided deep learning for dynamical systems: A survey,” arXiv preprint arXiv:2107.01272 , 2021

    2021 arXiv

  36. [44]

    Theory- training deep neural networks for an alloy solidification benchmark problem,

    M. Torabi Rad, A. Viardin, G. Schmitz, and M. Apel, “Theory- training deep neural networks for an alloy solidification benchmark problem,” Computational Materials Science , vol. 180, p. 109687, 2020. [Online]. Available: https://www.sciencedirect.com/science/article/pii/ S0927...

    2020

  37. [45]

    Hamiltonian neural net- works,

    S. Greydanus, M. Dzamba, and J. Yosinski, “Hamiltonian neural net- works,” Advances in neural information processing systems , vol. 32, 2019

    2019

  38. [46]

    Deep hidden physics models: Deep learning of nonlinear partial differential equations,

    M. Raissi, “Deep hidden physics models: Deep learning of nonlinear partial differential equations,” Journal of Machine Learning Research , vol. 19, no. 25, pp. 1–24, 2018

    2018

  39. [47]

    Symplectic ode-net: Learning hamiltonian dynamics with control,

    Y . D. Zhong, B. Dey, and A. Chakraborty, “Symplectic ode-net: Learning hamiltonian dynamics with control,” arXiv preprint arXiv:1909.12077 , 2019

    1909 arXiv

  40. [48]

    Dynamic modeling of robotic manipulator via an augmented deep lagrangian network,

    S. Wu, Z. Li, W. Chen, and F. Sun, “Dynamic modeling of robotic manipulator via an augmented deep lagrangian network,” Tsinghua Science and Technology, vol. 29, no. 5, pp. 1604–1614, 2024

    2024

  41. [49]

    Dynamics consensus between centroidal and whole-body models for locomotion of legged robots,

    R. Budhiraja, J. Carpentier, and N. Mansard, “Dynamics consensus between centroidal and whole-body models for locomotion of legged robots,” in 2019 International Conference on Robotics and Automation (ICRA). IEEE, 2019, pp. 6727–6733

    2019

  42. [50]

    Mahapatra, S

    A. Mahapatra, S. S. Roy, and D. K. Pratihar, Multi-body Dynamic Modeling of Multi-legged Robots . Springer Nature, 2020

    2020

  43. [51]

    Dynamic analysis tool for legged robots,

    F. Ouezdou, O. Bruneau, and J. Guinot, “Dynamic analysis tool for legged robots,” Multibody System Dynamics, vol. 2, no. 4, pp. 369–391, 1998

    1998

  44. [52]

    Modular framework for dynamic modeling and analyses of legged robots,

    S. Shah, S. Saha, and J. Dutt, “Modular framework for dynamic modeling and analyses of legged robots,” Mechanism and Machine Theory, vol. 49, pp. 234–255, 2012

    2012

  45. [53]

    Dynamic modeling, analysis, and comparative study of a quadruped with bio-inspired robotic tails,

    Y . Liu and P. Ben-Tzvi, “Dynamic modeling, analysis, and comparative study of a quadruped with bio-inspired robotic tails,” Multibody System Dynamics, vol. 51, no. 2, pp. 195–219, 2021

    2021

  46. [54]

    Trajec- tory optimization through contacts and automatic gait discovery for quadrupeds,

    M. Neunert, F. Farshidian, A. W. Winkler, and J. Buchli, “Trajec- tory optimization through contacts and automatic gait discovery for quadrupeds,” IEEE Robotics and Automation Letters , vol. 2, no. 3, pp. 1502–1509, 2017

    2017

  47. [55]

    Whole-body nonlinear model predictive control through contacts for quadrupeds,

    M. Neunert, M. St ¨auble, M. Giftthaler, C. D. Bellicoso, J. Carius, C. Gehring, M. Hutter, and J. Buchli, “Whole-body nonlinear model predictive control through contacts for quadrupeds,” IEEE Robotics and Automation Letters, vol. 3, no. 3, pp. 1458–1465, 2018

    2018

  48. [56]

    Dynamic locomotion and whole-body control for quadrupedal robots,

    C. D. Bellicoso, F. Jenelten, P. Fankhauser, C. Gehring, J. Hwangbo, and M. Hutter, “Dynamic locomotion and whole-body control for quadrupedal robots,” in 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) . IEEE, 2017, pp. 3359–3365

    2017

  49. [57]

    Dynamic formulation using the decoupled natural orthogonal complement (denoc),

    P. V . Nandihal, A. Mohan, S. K. Saha, P. V . Nandihal, A. Mohan, and S. K. Saha, “Dynamic formulation using the decoupled natural orthogonal complement (denoc),” Dynamics of Rigid-Flexible Robots and Multibody Systems , pp. 29–42, 2022

    2022

  50. [58]

    Nonlinear mpc for free-flying space manipulator,

    C. Sirois, “Nonlinear mpc for free-flying space manipulator,” TESI DI LAUREA MAGISTRALE, Politecnico di Milano , 2021

    2021

  51. [59]

    Geometric methods and formulations in computational multibody system dynamics,

    A. M ¨uller and Z. Terze, “Geometric methods and formulations in computational multibody system dynamics,” Acta mechanica, vol. 227, pp. 3327–3350, 2016

    2016

  52. [60]

    D. A. Lind, D. Lind, and B. Marcus, An introduction to symbolic dynamics and coding . Cambridge university press, 2021

    2021

  53. [61]

    Constrained multibody dynamics with python: From symbolic equation generation to publication,

    G. Gede, D. L. Peterson, A. S. Nanjangud, J. K. Moore, and M. Hub- bard, “Constrained multibody dynamics with python: From symbolic equation generation to publication,” in International Design Engineering Technical Conferences and Computers and Information in Engineering Confe...

    2013

  54. [62]

    Symbolic derivation of dynamic equations of motion for robot manipulators using piogram symbolic method,

    P.-Y . Cheng, C.-I. Weng, and C.-K. Chen, “Symbolic derivation of dynamic equations of motion for robot manipulators using piogram symbolic method,” IEEE Journal on Robotics and Automation , vol. 4, no. 6, pp. 599–609, 1988

    1988

  55. [63]

    A fully symbolic generation of the equations of motion of multibody systems containing flexible beams,

    P. Fisette, D. Johnson, and J.-C. Samin, “A fully symbolic generation of the equations of motion of multibody systems containing flexible beams,” Computer Methods in Applied Mechanics and Engineering, vol. 142, no. 1-2, pp. 123–152, 1997

    1997

  56. [64]

    Symbolic genetic algorithm for discovering open-form partial differential equations (sga- pde),

    Y . Chen, Y . Luo, Q. Liu, H. Xu, and D. Zhang, “Symbolic genetic algorithm for discovering open-form partial differential equations (sga- pde),” Physical Review Research, vol. 4, no. 2, p. 023174, 2022

    2022

  57. [65]

    R. S. Ball, A Treatise on the Theory of Screws . Cambridge university press, 1998

    1998

  58. [66]

    Robotic manipulators and the product of exponentials formula,

    R. W. Brockett, “Robotic manipulators and the product of exponentials formula,” in Mathematical theory of networks and systems . Springer, 1984, pp. 120–129

    1984

  59. [67]

    Murray, Z

    R. Murray, Z. Lie, and S. Sastry, A Mathematical Introduction to Robotic Manipulation. CRC Press, 1994

    1994

  60. [68]

    Computational aspects of the product-of-exponentials for- mula for robot kinematics,

    F. C. Park, “Computational aspects of the product-of-exponentials for- mula for robot kinematics,” IEEE Transactions on Automatic Control , vol. 39, no. 3, pp. 643–647, 1994

    1994

  61. [69]

    Robotic machining: A review of recent progress,

    S. H. Kim, E. Nam, T. I. Ha, S.-H. Hwang, J. H. Lee, S.-H. Park, and B.- K. Min, “Robotic machining: A review of recent progress,” International Journal of Precision Engineering and Manufacturing, vol. 20, pp. 1629– 1642, 2019

    2019

  62. [70]

    Stramigioli, Modeling and IPC control of interactive mechanical systems—A coordinate-free approach

    S. Stramigioli, Modeling and IPC control of interactive mechanical systems—A coordinate-free approach. Springer, 2001

    2001

  63. [71]

    Singularity-free lagrange-poincare equations on lie groups for vehicle-manipulator systems,

    B. M. Moghaddam and R. Chhabra, “Singularity-free lagrange-poincare equations on lie groups for vehicle-manipulator systems,” IEEE Trans- actions on Robotics , 2023

    2023

  64. [72]

    The reduced euler-lagrange equations,

    J. E. Marsden and J. Scheurle, “The reduced euler-lagrange equations,” Fields Institute Comm , vol. 1, no. 6, pp. 139–164, 1993

    1993

  65. [73]

    Singularity-free lagrange-poincar ´e equations on lie groups for vehicle-manipulator systems,

    B. M. Moghaddam and R. Chhabra, “Singularity-free lagrange-poincar ´e equations on lie groups for vehicle-manipulator systems,” IEEE Trans- actions on Robotics , vol. 40, pp. 1393–1409, 2024

    2024

  66. [74]

    The reduced euler–lagrange equations,

    J. M. Scheurle and J, “The reduced euler–lagrange equations,” Fields Institute Comm, vol. 1, pp. 139–164, 1993

    1993

  67. [75]

    Nonlinear control of mechanical systems: A lagrangian perspective,

    R. M. Murray, “Nonlinear control of mechanical systems: A lagrangian perspective,” Annual Reviews in Control , vol. 21, pp. 31–42, 1997

    1997

  68. [76]

    Lagrangian dynamics of open multi- body systems with generalized holonomic and nonholonomic joints,

    V . Duindam and S. Stramigioli, “Lagrangian dynamics of open multi- body systems with generalized holonomic and nonholonomic joints,” in 2007 IEEE/RSJ International Conference on Intelligent Robots and Systems. IEEE, 2007, pp. 3342–3347

    2007

  69. [77]

    A hamiltonian formulation of the dynamics of spatial mechanisms using lie groups and screw theory,

    S. Stramigioli, B. Maschke, and C. Bidard, “A hamiltonian formulation of the dynamics of spatial mechanisms using lie groups and screw theory,” in Proc. Symposium Commemorating the Legacy, Work and Life of Sir RS Ball, J. Duffy and H. Lipkin organizers , 2000, pp. 9–11

    2000

  70. [78]

    Corrections to “singularity- free dynamic equations of open-chain mechanisms with general holo- nomic and nonholonomic joints

    P. J. From, V . Duindam, and S. Stramigioli, “Corrections to “singularity- free dynamic equations of open-chain mechanisms with general holo- nomic and nonholonomic joints”,” IEEE transactions on robotics , vol. 28, no. 6, pp. 1431–1432, 2012

    2012

  71. [79]

    Singularity-free dynamic equations of vehicle–manipulator systems,

    P. J. From, V . Duindam, K. Y . Pettersen, J. T. Gravdahl, and S. Sastry, “Singularity-free dynamic equations of vehicle–manipulator systems,” Simulation Modelling Practice and Theory , vol. 18, no. 6, pp. 712–731, 2010

    2010

  72. [80]

    Singularity-free dynamic equations of auv-manipulator systems,

    P. J. From, K. Y . Pettersen, and J. T. Gravdahl, “Singularity-free dynamic equations of auv-manipulator systems,” IFAC Proceedings Volumes , vol. 43, no. 16, pp. 31–36, 2010

    2010

  73. [81]

    A singularity free formulation of the dynamically equivalent manipulator mapping of space manipulators,

    ——, “A singularity free formulation of the dynamically equivalent manipulator mapping of space manipulators,” AIAA SPACE Conference & Exposition, pp. 1–10, 2011

    2011

  74. [82]

    An error-state model predictive control on connected matrix lie groups for legged robot control,

    S. Teng, D. Chen, W. Clark, and M. Ghaffari, “An error-state model predictive control on connected matrix lie groups for legged robot control,” in 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) . IEEE, 2022, pp. 8850–8857

    2022

  75. [83]

    Geometric control and learning for dynamic legged robots,

    A. Siravuru, “Geometric control and learning for dynamic legged robots,” Ph.D. dissertation, Carnegie Mellon University, 2020

    2020

  76. [84]

    M. M. Rohde, J. Crawford, M. Toschlog, K. D. Iagnemma, G. Kewlani, C. L. Cummins, R. A. Jones, and D. A. Horner, “An interactive physics-based unmanned ground vehicle simulator leveraging open source gaming technology: progress in the development and application of the virtual...

    2009

  77. [85]

    High-fidelity yet fast dynamic models of wheeled mobile robots,

    N. Seegmiller and A. Kelly, “High-fidelity yet fast dynamic models of wheeled mobile robots,” IEEE Transactions on robotics , vol. 32, no. 3, pp. 614–625, 2016

    2016

  78. [86]

    Open dynamics engine,

    R. Smith et al., “Open dynamics engine,” 2005

    2005

  79. [87]

    Rigid body dynamics library,

    M. Felis., “Rigid body dynamics library,” Available online at: http://rbdl.bitbucket.org/

  80. [88]

    Opensymoro: An open-source software package for symbolic modelling of robots,

    W. Khalil, A. Vijayalingam, B. Khomutenko, I. Mukhanov, P. Lemoine, and G. Ecorchard, “Opensymoro: An open-source software package for symbolic modelling of robots,” in 2014 IEEE/ASME International Conference on Advanced Intelligent Mechatronics . IEEE, 2014, pp. 1206–1211

    2014

  81. [89]

    Robots that can adapt like animals,

    A. Cully, J. Clune, D. Tarapore, and J.-B. Mouret, “Robots that can adapt like animals,” Nature, vol. 521, no. 7553, pp. 503–507, 2015

    2015

  82. [90]

    Resilient machines through continuous self-modeling,

    J. Bongard, V . Zykov, and H. Lipson, “Resilient machines through continuous self-modeling,” Science, vol. 314, no. 5802, pp. 1118–1121, 2006

    2006

  83. [91]

    A generalized exponential formula for forward and differential kinematics of open-chain multibody systems,

    R. Chhabra and M. Emami, “A generalized exponential formula for forward and differential kinematics of open-chain multibody systems,” Mechanism and Machine Theory , vol. 73, pp. 61–75, 2014

    2014

  84. [92]

    Singularity-free whole-body dynamical equations of legged robots for damage simulation,

    S. Farghdani and R. Chhabra, “Singularity-free whole-body dynamical equations of legged robots for damage simulation,” in AIAA SCITECH 2024 Forum, 2024, p. 1017

    2024

  85. [93]

    A numeric derivation for fast regressive modeling of manipulator dynamics,

    S. Lloyd, R. Irani, and M. Ahmadi, “A numeric derivation for fast regressive modeling of manipulator dynamics,”Mechanism and Machine Theory, vol. 156, p. 104149, 2021

    2021

  86. [94]

    Available online at: https://www.hiwonder.com/products/jethexa (Accessed March, 2024)

    Hiwonder JetHexa ROS Hexapod Robot Kit Powered by Jetson Nano Boston Dynamics.; 2022. Available online at: https://www.hiwonder.com/products/jethexa (Accessed March, 2024). APPENDIX A PROOF OF LEMMA 3 Proof. GivenMij from (30) and knowing that the only vari- able part of the e...

    2022

Pith tools

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