pith. sign in

arxiv: 2604.12274 · v1 · submitted 2026-04-14 · 💻 cs.RO

Asymptotically Stable Gait Generation and Instantaneous Walkability Determination for Planar Almost Linear Biped with Knees

Fumihiko Asano , Ning Lei , Taiki Sedoguchi This is my paper

Pith reviewed 2026-05-10 16:14 UTC · model grok-4.3

classification 💻 cs.RO
keywords bipedal gaitknee jointslinear approximationwalkability determinationplanar robotstable gait generationequations of motioncontrol system
0
0 comments X

The pith

A linear approximation of gravity in the knee biped model enables instantaneous calculation of future states and walkability.

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

The paper shows that bipeds with knees can have their dynamics simplified by linearizing the gravity term around the thigh angle after reducing to a 3-DOF model. This avoids numerical integration for predicting states or checking if a step is walkable. The approach builds on constant inertia and no Coriolis terms in the original almost-linear design, allowing control inputs and discretization for fast iterative checks. A sympathetic reader would care because real-time gait planning and stability verification become feasible even on modest hardware.

Core claim

By deriving the equations of motion for a planar 6-DOF biped with knees, reducing to 3-DOF, and approximating the gravity term linearly around the thigh frame angle, the model permits state-space realization and discretization. This allows iterative calculation for instantaneous walkability determination while generating asymptotically stable gaits on level ground.

What carries the argument

The linearized 3-DOF model from Taylor expansion of the gravity term around the thigh frame angle, which carries the argument by enabling fast state prediction without integration.

If this is right

  • Stable bipedal gaits can be numerically generated using the control system on the linearized model.
  • Future states and walkability can be determined in negligible computation time.
  • Accuracy of the approximation depends on the knee flexion angle and chosen expansion point.
  • Small steps introduce issues for the linear approximation when descending.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The method could extend to real-time control on physical robots by avoiding heavy simulation.
  • Choosing the expansion point optimally might improve accuracy for varied terrains beyond small steps.
  • Similar linearizations could apply to other joint configurations if inertia remains constant.

Load-bearing premise

The linear approximation of the gravity term around the thigh frame angle stays accurate enough for the knee flexion angles that occur during the gait, even on small steps.

What would settle it

Run a full nonlinear simulation of a generated gait and compare the predicted future states from the linearized discrete model; if the positions or velocities diverge significantly within one step, the claim fails.

Figures

Figures reproduced from arXiv: 2604.12274 by Fumihiko Asano, Ning Lei, Taiki Sedoguchi.

Figure 1
Figure 1. Figure 1: Model of planar, 6-DOF, almost linear biped with knee [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Simulation results of asymptotically stable walkin [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Evolution of gait descriptors in asymptotically stab [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 5
Figure 5. Figure 5: Evolution of gait descriptors during generated gait [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗
read the original abstract

A class of planar bipedal robots with unique mechanical properties has been proposed, where all links are balanced around the hip joint, preventing natural swinging motion due to gravity. A common property of their equations of motion is that the inertia matrix is a constant matrix, there are no nonlinear velocity terms, and the gravity term contains simple nonlinear terms. By performing a Taylor expansion of the gravity term and making a linear approximation, it is easy to derive a linearized model, and calculations for future states or walkability determination can be performed instantaneously without the need for numerical integration. This paper extends the method to a planar biped robot model with knees. First, we derive the equations of motion, constraint conditions, and inelastic collisions for a planar 6-DOF biped robot, design its control system, and numerically generate a stable bipedal gait on a horizontal plane. Next, we reduce the equations of motion to a 3-DOF model, and derive a linearized model by approximating the gravity term as linear around the expansion point for the thigh frame angle. Through numerical simulations, we demonstrate that calculations for future states and walkability determination can be completed in negligible time. By applying control inputs to the obtained model, performing state-space realization, and then discretizing it, instantaneous walkability determination through iterative calculation becomes possible. Through detailed gait analysis, we discuss how the knee joint flexion angle and the expansion point affect the accuracy of the linear approximation, and the issues that arise when descending a small step.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript proposes an approach for asymptotically stable gait generation and fast walkability determination in a planar biped with knees exhibiting almost-linear dynamics. The authors derive the full 6-DOF equations of motion, reduce them to a 3-DOF model, perform a linear approximation of the gravity vector around a fixed thigh-frame angle, obtain a state-space realization, discretize the system, and use iterative forward simulation for instantaneous walkability checks. Numerical examples demonstrate stable gait generation on level ground and discuss the impact of knee flexion on approximation accuracy, including for small steps.

Significance. If the linear gravity approximation proves sufficiently accurate over the operating range of knee angles, the method offers a computationally efficient alternative to full nonlinear integration for real-time locomotion planning. The numerical results on gait stability and computation time are promising, but the absence of quantitative validation metrics against the nonlinear model weakens the assessment of practical utility.

major comments (2)
  1. Linearized model derivation (following reduction to 3-DOF): the first-order Taylor expansion is applied only to the thigh-frame angle while gravity terms depend on both thigh and knee angles; without an explicit remainder bound or sensitivity analysis over the observed knee-flexion range (especially for small steps), the accuracy of the discrete linear model for walkability decisions is not rigorously established.
  2. Numerical simulations section: stable gaits are generated and computation times reported as negligible, yet no quantitative error metrics (e.g., trajectory L2 norms, state-prediction discrepancies, or walkability-decision mismatch rates) are given comparing the discretized linear model to forward integration of the original nonlinear 6-DOF system, which is load-bearing for the central claim of reliable instantaneous determination.
minor comments (2)
  1. Abstract and gait-analysis discussion: the choice of fixed expansion point for the thigh angle is not justified with a selection criterion or adaptation strategy, and its interaction with varying knee flexion should be clarified for reproducibility.
  2. The manuscript would benefit from a table summarizing computation times, approximation errors (if any), and knee-angle ranges across the reported gaits.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We are grateful to the referee for the constructive and detailed review of our manuscript. The comments correctly identify areas where additional rigor would strengthen the presentation of the linear approximation and its validation. We address each major comment below and will incorporate revisions to provide the requested analysis and metrics.

read point-by-point responses

  1. Referee: Linearized model derivation (following reduction to 3-DOF): the first-order Taylor expansion is applied only to the thigh-frame angle while gravity terms depend on both thigh and knee angles; without an explicit remainder bound or sensitivity analysis over the observed knee-flexion range (especially for small steps), the accuracy of the discrete linear model for walkability decisions is not rigorously established.

    Authors: In the reduced 3-DOF model the knee angle enters the gravity vector as a configuration variable, and the first-order expansion is performed with respect to the thigh angle at a chosen operating point. The manuscript already contains a discussion of how knee flexion and the expansion point affect approximation accuracy, including explicit remarks on small-step cases. We agree, however, that an explicit remainder bound and systematic sensitivity analysis are absent. In the revision we will add a dedicated paragraph deriving a Lagrange remainder estimate for the gravity term over the observed knee-angle range and include sensitivity plots of the approximation error versus knee flexion. revision: yes

  2. Referee: Numerical simulations section: stable gaits are generated and computation times reported as negligible, yet no quantitative error metrics (e.g., trajectory L2 norms, state-prediction discrepancies, or walkability-decision mismatch rates) are given comparing the discretized linear model to forward integration of the original nonlinear 6-DOF system, which is load-bearing for the central claim of reliable instantaneous determination.

    Authors: The numerical section demonstrates stable gaits, negligible run times, and qualitative effects of knee flexion on accuracy, but does not supply direct quantitative comparisons against the full nonlinear 6-DOF integration. We accept that such metrics are necessary to substantiate the reliability claim for instantaneous walkability determination. In the revised manuscript we will add a new table and accompanying text reporting L2 trajectory errors, maximum state-prediction discrepancies over multiple steps, and walkability-decision agreement rates between the discretized linear model and nonlinear forward integration for the level-ground and small-step cases already presented. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivations from standard rigid-body mechanics with explicit approximation

full rationale

The paper starts from the balanced-link property and derives the 6-DOF equations of motion, constraint conditions, and collision models using standard Lagrangian mechanics. It then reduces to a 3-DOF model and applies an explicit first-order Taylor expansion to linearize only the gravity vector around a chosen thigh-frame angle. This linearized model is realized in state-space form, discretized, and iterated for walkability; the approximation is presented as such and its accuracy is assessed via separate numerical gait simulations rather than being tautologically defined by the walkability output. No parameters are fitted to data and relabeled as predictions, no uniqueness theorem is imported from self-citation to force the result, and the central claim (instantaneous walkability via the discrete linear model) follows directly from the stated approximation without reducing to its own inputs by construction.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The method depends on the mechanical property that produces constant inertia and absent Coriolis terms, plus the validity of the chosen linearization point; these are domain assumptions rather than derived results.

free parameters (1)
  • expansion point for thigh frame angle
    Angle chosen for Taylor expansion of gravity term; directly affects linear model accuracy as noted in the gait analysis discussion.
axioms (1)
  • domain assumption Inertia matrix is constant and nonlinear velocity terms are absent because all links are balanced around the hip joint
    Stated as a common property of the robot class that enables the almost-linear model.

pith-pipeline@v0.9.0 · 5583 in / 1251 out tokens · 68305 ms · 2026-05-10T16:14:18.038292+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

40 extracted references · 40 canonical work pages

  1. [1]

    11em plus .33em minus .07em @technote 4000 4000 100 4000 4000 500 `\.=1000 = #1 #1 #1 0pt [0pt][0pt] #1 * \| ** #1 \@IEEEauthorblockNstyle \@IEEEauthorblockAstyle \@IEEEauthordefaulttextstyle \@IEEEauthorblockconfadjspace -0.25em \@IEEEauthorblockNtopspace 0.0ex \@IEEEauthorblockAtopspace 0.0ex \@IEEEauthorblockNinterlinespace 2.6ex \@IEEEauthorblockAinte...

    work page 2026

  2. [2]

    McGeer, ``Passive dynamic walking,'' Int

    T. McGeer, ``Passive dynamic walking,'' Int. J. of Robotics Research , Vol. 9, Iss. 2, pp. 62--82, 1990

    work page 1990

  3. [3]

    McGeer, ``Passive walking with knees,'' Proc

    T. McGeer, ``Passive walking with knees,'' Proc. of the IEEE Int. Conf. on Robotics and Automation , Vol. 3, pp. 1640--1645, 1990

    work page 1990

  4. [4]

    J. W. Grizzle, G. Abba and F. Plestan. ``Asymptotically stable walking for biped robots: Analysis via systems with impulse effects,'' IEEE Trans. on Automatic Control , Vol. 46, Iss. 1, pp. 51--64, 2001

    work page 2001

  5. [5]

    E. R. Westervelt, J. W. Grizzle, C. Chevallereau, J. H. Choi, and B. Morris, Feedback Control of Dynamic Bipedal Robot Locomotion , CRC Press, 2007

    work page 2007

  6. [6]

    Kiefer and B

    J. Kiefer and B. Ramesh, ``Walking viability and gait synthesis for a novel class of dynamically-simple bipeds.'' Informatica , Vol. 17, Iss. 2, pp. 145--155, 1993

    work page 1993

  7. [7]

    M. W. Spong, R. Lozano and R. Mahony, ``An almost linear biped,'' Proc. of the 39th IEEE Conf. on Decision and Control , pp. 4803--4808, 2000

    work page 2000

  8. [8]

    S. K. Agrawal and A. Fattah, ``Motion control of a novel planar biped with nearly linear dynamics,'' IEEE/ASME Trans. on Mechatronics , Vol. 11, Iss. 2, pp. 162--168, 2006

    work page 2006

  9. [9]

    Sangwan and S

    V. Sangwan and S. K. Agrawal, ``Differentially flat design of bipeds ensuring limit cycles,'' IEEE/ASME Trans. on Mechatronics , Vol. 14, Iss. 6, pp. 647--657, 2009

    work page 2009

  10. [10]

    Asano, ``A novel generation method for underactuated bipedal gait with landing position control of swing leg based on property of zero dynamics,'' Proc

    F. Asano, ``A novel generation method for underactuated bipedal gait with landing position control of swing leg based on property of zero dynamics,'' Proc. of the IEEE Int. Conf. on Advanced Intelligent Mechatronics , pp. 1184-1189, 2015

    work page 2015

  11. [11]

    Li and Q

    J. Li and Q. Nguyen, ``Dynamic walking of bipedal robots on uneven stepping stones via adaptive-frequency MPC,'' IEEE Control Systems Letters , Vol. 7, pp. 1279--1284, 2023

    work page 2023

  12. [12]

    Jenelten, R

    F. Jenelten, R. Grandia, F. Farshidian and M. Hutter, ``TAMOLS: Terrain-aware motion optimization for legged systems,'' IEEE Trans. on Robotics , Vol. 38, Iss. 6, pp. 3395--3413, 2022

    work page 2022

  13. [13]

    Asano, ``Stability analysis of underactuated compass gait based on linearization of motion,'' Multibody System Dynamics , Vol

    F. Asano, ``Stability analysis of underactuated compass gait based on linearization of motion,'' Multibody System Dynamics , Vol. 33, Iss. 1, pp. 93--111, 2015

    work page 2015

  14. [14]

    Asano, ``Fully analytical solution to discrete behavior of hybrid zero dynamics in limit cycle walking with constraint on impact posture,'' Multibody System Dynamics , Vol

    F. Asano, ``Fully analytical solution to discrete behavior of hybrid zero dynamics in limit cycle walking with constraint on impact posture,'' Multibody System Dynamics , Vol. 35, Iss. 2, pp. 191--213, 2015

    work page 2015

  15. [15]

    M. J. Coleman, ``Dynamics and stability of a rimless spoked wheel: a simple 2D system with impacts,'' Dynamical Systems , Vol. 25, Iss. 2, pp. 215--238, 2010

    work page 2010

  16. [16]

    M. J. Coleman, A. Chatterjee and A. Ruina, ``Motions of a rimless spoked wheel: a simple three-dimensional system with impacts,'' Dynamics and Stability of Systems , Vol. 12, Iss. 3, pp. 139--159, 1997

    work page 1997

  17. [17]

    Asano, ``Stability principle underlying passive dynamic walking of rimless wheel,'' Proc

    F. Asano, ``Stability principle underlying passive dynamic walking of rimless wheel,'' Proc. of the IEEE Int. Conf. on Control Applications , pp. 1039--1044, 2012

    work page 2012

  18. [18]

    T. G. McGee and M. W. Spong, ``Trajectory planning and control of a novel walking biped,'' Proc. of the IEEE Int. Conf. on Control Applications , pp. 1099--1104, 2001

    work page 2001

  19. [19]

    Asano and Y

    F. Asano and Y. Zheng, ``High-speed and energy-efficient collisionless walking of underactuated rimless wheel,'' Artificial Life and Robotics , Vol. 23, Iss. 4, pp. 523--531, 2018

    work page 2018

  20. [20]

    Asano, T

    F. Asano, T. Sedoguchi and C. Yan, ``Generation of steady wheel gait for planar X-shaped walker with reaction wheel,'' Proc. of the 2024 IEEE International Conference on Robotics and Automation , pp. 5766--5772, 2024

    work page 2024

  21. [21]

    Asano and T

    F. Asano and T. Sedoguchi, ``Instantaneous walkability determination method for almost linear passive dynamic walkers with non-trivial limit cycle stability,'' Proc. of the IEEE/RSJ Int. Conf. on Intelligent Robots and Systems , 2025

    work page 2025

  22. [22]

    Hanazawa, Y

    Y. Hanazawa, Y. Uchino, and S. Sagara. ``Development of a rimless wheel robot with telescopic legs for step adaptability.'' Artificial Life and Robotics , Vol. 29, Iss. 2, pp. 349--357, 2024

    work page 2024

  23. [23]

    Asano and Y

    F. Asano and Y. He, ``Water surface walking of planar 6-DOF legged robot based on control of center of mass,'' Int. J. of Dynamics and Control , Vol. 13, article number 153, 2025

    work page 2025

  24. [24]

    Sanchez and P

    S. Sanchez and P. A. Bhounsule. ``Design, modeling, and control of a differential drive rimless wheel that can move straight and turn.'' Automation , Vol. 2, Iss. 3, pp. 98--115, 2021

    work page 2021

  25. [25]

    Asano and J

    F. Asano and J. Kawamoto, ``Modeling and analysis of passive viscoelastic-legged rimless wheel that generates measurable period of double-limb support,'' Multibody System Dynamics , Vol. 31, Iss. 2, pp. 111--126, 2014

    work page 2014

  26. [26]

    ``Walking experiments of small and lightweight rimless wheel robot.'' Artificial Life and Robotics 27.4 (2022): 706-713

    Hanazawa, Yuta, Haruka Nishinami, and Shinichi Sagara. ``Walking experiments of small and lightweight rimless wheel robot.'' Artificial Life and Robotics 27.4 (2022): 706-713

    work page 2022

  27. [27]

    Shankar and J

    K. Shankar and J. W. Burdick. ``Motion planning and control for a tethered, rimless wheel differential drive vehicle.'' Proc. of the IEEE/RSJ Int. Conf. on Intelligent Robots and Systems , 2013

    work page 2013

  28. [28]

    Gait Analysis and Control of Rimless Wheel Rovers for Planetary Exploration

    SCHLEE, Dennis, et al. Gait Analysis and Control of Rimless Wheel Rovers for Planetary Exploration. In: 2024 International Conference on Space Robotics (iSpaRo). IEEE, 2024. p. 139-144

    work page 2024

  29. [29]

    Asano, Y

    F. Asano, Y. Xiang, Y. Zheng and C. Yan, ``Modeling, analysis and activation of planar viscoelastically-combined rimless wheels,'' Proc. of the IEEE/RSJ Int. Conf. on Intelligent Robots and Systems , pp. 12957--12962, 2022

    work page 2022

  30. [30]

    Zheng, F

    Y. Zheng, F. Asano, C. Yan, L. Li and I. T. Tokuda, ``Tensegrity-based legged robot generates passive walking, skipping, and crawling gaits in accordance with environment,'' IEEE/ASME Trans. on Mechatronics , Published in Early Access on Xplore, Jan. 10, 2025

    work page 2025

  31. [31]

    Asano, T

    F. Asano, T. Sedoguchi and I. Tokuda, ``Mathematical modeling and rolling motion generation of planar seven-link robot that forms passive closed and active open chains,'' Proc. of the IEEE Int. Conf. on Robotics and Automation , 2025. (to apper)

    work page 2025

  32. [32]

    Mori and T

    O. Mori and T. Omata, ``Coupling of two 2-link robots with a passive joint for reconfigurable planar parallel robot,'' Proc. of the IEEE Int. Conf. on Robotics and Automation , pp. 4120--4125, 2002

    work page 2002

  33. [33]

    W. H. Lee and A. C. Sanderson, ``Dynamic rolling locomotion and control of modular robots,'' IEEE Trans. on Robotics and Automation , pp. 32--41, Vol. 18, Iss. 1, 2002

    work page 2002

  34. [34]

    Yamawaki, O

    T. Yamawaki, O. Mori and T. Omata, ``Nonholonomic dynamic rolling control of reconfigurable 5R closed kinematic chain robot with passive joints,'' Proc. of the IEEE Int. Conf. on Robotics and Automation , pp. 4054--4059, 2003

    work page 2003

  35. [35]

    Matsuda and S

    T. Matsuda and S. Murata, ``Stiffness distribution control - locomotion of closed link robot with mechanical softness,'' Proc. of the IEEE Int. Conf. on Robotics and Automation , pp. 1491--1498, 2006

    work page 2006

  36. [36]

    Xiangzhi, T

    W. Xiangzhi, T. Yaobin and W. Shanshan, ``Design and locomotion analysis of a novel modular rolling robot,'' Mechanism and Machine Theory , Vol. 133, pp. 23--43, 2019

    work page 2019

  37. [37]

    Y. Hao, Y. Tian, J. Wu, Y. Li and Y. Yao, ``Design and locomotion analysis of two kinds of rolling expandable mobile linkages with a single degree of freedom,'' Frontiers of Mechanical Engineering , Vol. 15. pp. 365--373, 2020

    work page 2020

  38. [38]

    Y. Li, Z. Wang, Y. Xu, J. S. Dai, Z. Zhao and Y. Yao, ``A deformable tetrahedron rolling mechanism (DTRM) based on URU branch,'' Mechanism and Machine Theory , Vol. 153, Article 104000, 2020

    work page 2020

  39. [39]

    Paskarbeit, S

    J. Paskarbeit, S. Beyer, M. Engel, A. Gucze, J. Schr\" o der and A. Schneider, ''Ourobot--A sensorized closed-kinematic-chain robot for shape-adaptive rolling in rough terrain,'' Robotics and Autonomous Systems , Vol. 140, 2021

    work page 2021

  40. [40]

    E. R. Westervelt, J. W. Grizzle and D. E. Koditschek. ``Zero dynamics of planar biped walkers with one degree of under actuation,'' Vol. 35, Iss. 1, pp. 551--556, 2002. 0.8pt eqnarray* &=& [ array cccccc 0 & 0 & 0 & 1 & 0 & 0 \\ 0 & 0 & 0 & 0 & 1 & 0 \\ 0 & 0 & 0 & 0 & 0 & 1 \\ ^2 & 0 & 0 & 0 & 0 & 0 \\ ^2 & 0 & 0 & 0 & 0 & 0 \\ ^2 & 0 & 0 & 0 & 0 & 0 arr...

    work page 2002