Pith. sign in

REVIEW 1 major objections 5 minor 1 cited by

Modular Robot Control with Motor Primitives

T0 review · 1 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read Four basic impedance modules satisfy the two properties the authors define as modularity—independence and closure of stability—removing inverse kinematics, preserving passivity in contact, and traversing kinematic singularities.

desk verdict Solid synthesis with a genuinely useful modularity definition and nice demos, but the time-varying passivity proof has a real gap at zero velocity. read the letter →

arxiv 2505.10694 v1 pith:HHSXCJJ4 submitted 2025-05-15 cs.RO

Moses C. Nah , Johannes Lachner , Neville Hogan This is my paper · ORCID
classification cs.RO
keywords MotorPrimitivesModularityDynamicMovementElementaryActionsKinematicSingularityRedundancyImpedanceControlPassivity
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 argues that robot control can be made genuinely modular, and that only four building blocks are needed: joint-space impedance, task-space position impedance, and task-space orientation impedance expressed either with rotation matrices or quaternions. Each module pairs a mechanical impedance with a virtual trajectory, and the paper defines modularity by two properties: independence, meaning any module can be modified without disturbing the others, and closure of stability, meaning any combination of individually stable modules remains stable, including during contact. The authors claim both properties hold for their four modules, and that this eliminates inverse kinematics, lets the robot pass through kinematic singularities, resolves kinematic redundancy without joint drift, and keeps the robot passive when touching passive environments. If the claim is right, complex behaviors such as shaking a cocktail or pouring from a bottle can be composed from learned or hand-specified primitives rather than solved as one monolithic control problem.

What carries the argument

The load-bearing object is the module: a pair $(Z, x_0)$ consisting of a mechanical impedance $Z$ and a virtual trajectory $x_0$ to which that impedance is connected, assembled through a Norton equivalent network model that separates the one-way 'information domain' of motion planning from the two-way 'physical domain' of force interaction. Two superposition principles carry the argument: virtual trajectories add linearly, $x_0(t)=\sum_i x_{0,i}(t)$, and mechanical impedances add, $Z=\sum_i Z_i$, so the total torque command is the gradient of a summed virtual elastic potential pulled back to joint space, $\tau_{in}=-\partial(U_q+U_p\circ h_p+U_r\circ h_r)/\partial q$. Stability is carried by the total energy as a storage function: with constant module parameters the derivative is $-\dot q^\top B_q \dot q \leq 0$ and LaSalle's invariance principle yields convergence to a local minimum of the potential, while for time-varying parameters passivity is asserted for 'sufficiently large' joint damping $B_q$ relative to $\partial U/\partial t$.

What would settle it

Re-run the paper's cocktail-shaking or circular-trajectory demonstrations on the same seven-joint arm while sweeping trajectory speed and joint damping $B_q$ downward, and measure energy exchange at a rigid instrumented contact such as a wall with a force/torque sensor. The central claim fails if the storage-function derivative $\dot V = -\dot q^\top B_q \dot q + \partial U/\partial t$ turns positive, because energy would then flow out of the robot into the environment, and for time-varying parameters the proof rests entirely on keeping that derivative non-positive; the same verdict can be reached without hardware by evaluating $\sup_t \partial U/\partial t$ against $\dot q^\top B_q \dot q$ on the recorded trajectories.

Watch

Extended reading notes

Core claim

The central claim is that a robot torque controller assembled from the four basic modules satisfies both properties the authors define as modularity: independence of modules and closure of stability. Independence is delivered by two superposition principles—virtual trajectories add linearly, and mechanical impedances add even though each is a nonlinear operator—so motions planned in joint space, task-space position, and task-space orientation can be planned separately and summed at the torque-command level. Closure of stability is delivered by an energy argument: the total virtual elastic potential pulled back to joint space, $U_q + U_p\circ h_p + U_r\circ h_r$, added to kinetic energy, forms a storage function whose time derivative is non-positive when joint damping is large enough, so passivity against passive environments follows. From this the paper derives task-space control without inverse kinematics, smooth passage through kinematic singularities, management of kinematic redundancy without joint drift, deliberate use of singular configurations to hold heavy external loads with low torque, and control of points off the robot's body.

Load-bearing premise

The load-bearing premise is that the friction-like damping in the robot's joints is always large enough, relative to how quickly the controller's virtual spring-energy landscape changes, to keep the robot from injecting energy into its surroundings—the paper says the damping must be 'sufficiently large' but gives no explicit bound, even though every demonstrated behavior uses time-varying trajectories.

Editorial extensions

If this is right

  • Torque-controlled robots can be commanded directly in task space because the controller uses only the forward kinematics map and the Jacobian transpose, never an inverse; kinematic singularities become passable passages rather than obstacles.
  • A kinematically redundant robot can run repeatable task-space motions without joint drift: adding the joint-space module with positive-definite stiffness removes the drift, at the cost of reduced tracking accuracy from task conflict.
  • Operating near a singularity lets a robot hold a heavy external load with very low torque—the paper demonstrates a 31 kg bookshelf held with joint torques below 10% of rated saturation—while preserving passivity.
  • Movements learned in different spaces, such as task-space position and orientation, can be learned separately by imitation and combined at the torque level, which the paper demonstrates by having the robot shake a cocktail from separately learned figure-eight and shaking primitives.
  • Because the virtual trajectory need not lie on the robot, the same controller stabilizes points outside the body, such as the tip of a bottle during pouring, with no change of algorithm.

Reading between the lines

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

  • The paper does not sweep module parameters or trajectory speeds, yet its independence claim invites exactly that stress test: combine a rhythmic and a discrete virtual trajectory over many phase offsets, amplitudes, and speeds, and check that neither module needs retuning and stability never degrades; a failure would pinpoint where independence actually breaks.
  • If closure of stability holds under the paper's unquantified 'sufficiently large damping' condition, the modules become a natural inductive bias for robot learning: a learner could tune impedances and trajectory weights with passivity as a permanent safety certificate, which is the constraint regime in which sample-efficient policy learning becomes tractable.
  • The paper's biological framing carries a falsifiable prediction it leaves implicit: if humans show measurable interference when combining rhythmic and discrete actions in one task, biological modularity would be weaker than the engineering independence defined here, which would bound the framework's descriptive reach.
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

1 major / 5 minor

Summary. The paper introduces a modular control framework for torque-controlled robots, combining Elementary Dynamic Actions (EDA) with Dynamic Movement Primitives (DMP). A module is defined as a mechanical impedance connected to a virtual trajectory, and four concrete modules are proposed: joint-space, task-space position, SO(3)-based orientation, and quaternion-based orientation. The central claim is that composing these modules through superposition of virtual trajectories and superposition of mechanical impedances yields modularity in the sense of two properties the authors define: independence of modules and closure of stability. The paper further claims that this architecture avoids inverse kinematics, traverses kinematic singularities seamlessly, resolves kinematic redundancy, preserves passivity under contact, and allows virtual trajectories outside the robot body. Validation is provided through MuJoCo simulations and KUKA iiwa14 experiments, with code and videos available.

Significance. The framework is a useful synthesis of EDA and DMP: the four modules are concrete, the constant-parameter stability proof in Section 3.4.2.1 is clean, and the demonstrations showcase genuinely relevant capabilities such as singularity traversal, redundancy management, contact, and object-centric manipulation. The explicit formulation of modularity via independence and closure is a valuable contribution, and the public code and reproducible demonstrations are strengths. However, the theoretical contribution is currently limited by the time-varying passivity claim in Section 3.4.2.2, which is not established and, as stated, is not pointwise satisfiable. If the authors supply a valid passivity condition or appropriately weaken the claim, the paper would be a solid contribution to modular robot control.

major comments (1)
  1. [Section 3.4.2.2] The passivity claim for time-varying module parameters is not established, and the stated sufficient condition cannot hold pointwise. The derivation gives dV/dt = -qdot^T B_q qdot + ∂U/∂t. Since ∂U/∂t is independent of qdot, a positive value of ∂U/∂t makes dV/dt > 0 at qdot = 0 no matter how large the joint damping B_q is chosen. For example, for a 1-DOF mass with tau = -K(q - q0(t)) - b qdot and q0(t) = v t, at q = 0 and qdot = 0 one has ∂U/∂t = K v^2 t > 0, so the storage function increases with zero external force and the controller is not passive with respect to the external port. A correct passivity statement would require dV/dt ≤ qdot^T tau_ext, which leads to the condition ∂U/∂t ≤ qdot^T B_q qdot, and this cannot be satisfied at qdot = 0 when ∂U/∂t > 0. Because all Section 4 demonstrations use time-varying virtual trajectories (minimum-jerk or DMP), the closure-of-stability/passivity guarantee claimed for those cases does not follow from the proof. Please either provide a valid sufficient condition, for example by bounding ∂U/∂t by a negative-definite function of qdot or by treating the virtual trajectory as an input with an appropriate supply rate, or restrict the passivity theorem to the constant-parameter case.
minor comments (5)
  1. [Throughout] The displayed equations are not numbered; please add equation numbers, since the text frequently refers to equations and future readers will need precise references.
  2. [Section 4] The phrase 'Exp[licit] TM-FRI Library' appears to be a bracket markup artifact; it should read 'Explicit TM-FRI Library'.
  3. [Section 5.6] The sentence 'the modules presented in this Thesis' should presumably read 'the modules presented in this paper', since the manuscript is a journal article rather than a thesis.
  4. [Section 4.1.4 and Figure 7] The quantitative claim about lower joint torques near singular configurations is supported by a single experiment; please consider adding repeated trials with error bars, or describe the result explicitly as a demonstration rather than as a validated quantitative comparison.
  5. [Figure 5] The '30% of the workspace' figure depends on the arbitrary singular-value threshold of 0.03; please state the sensitivity of this percentage to the threshold choice, or justify the threshold.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation found: the modularity claim rests on a self-contained Lyapunov argument, and the time-varying passivity gap is a correctness issue rather than a circular reduction.

full rationale

The claimed derivation chain is not circular. The four modules are explicit torque contributions (Eqs. 2, 4, 6, 8), and combining them is a linear summation of torques; the independence property follows from that additive structure rather than being assumed as the conclusion. The closure-of-stability proof for constant parameters is self-contained: with U = Uq + Up∘hp + Ur∘hr, the Lyapunov function V = ½ q̇ᵀM(q)q̇ + U(q) yields dV/dt = −q̇ᵀBq q̇ ≤ 0 using the skew-symmetry of Ṁ−2C and LaSalle's principle. No fitted parameter or target result is used as an input to that proof. The paper's prior work on EDA, the Norton equivalent network model, and earlier EDA/DMP comparisons supplies vocabulary and context, but the stability and passivity arguments in Section 3.4 do not reduce to those citations. The demonstrations in Section 4 are validations, not predictions from fitted quantities, and the superposition principles are stated as structural properties of the framework rather than as outputs of a fit. The main caveat is Section 3.4.2.2: the claim that passivity with time-varying parameters follows if Bq is 'sufficiently large relative to ∂U/∂t' is unsupported and fails pointwise at q̇ = 0 whenever ∂U/∂t > 0, because dV/dt = −q̇ᵀBq q̇ + ∂U/∂t is then positive regardless of Bq. That is a genuine correctness gap for the time-varying demonstrations, but it is not circularity: the flawed sufficient condition is not equivalent to its own conclusion and is not a fitted input renamed as a prediction.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central claim depends on standard control-theoretic results plus domain assumptions about ideal actuation and passive/admittance environments. The most fragile premise is the unquantified damping condition for time-varying passivity. The free parameters are controller gains and an analysis threshold, all chosen by hand without a systematic procedure.

free parameters (2)
  • Impedance gains (Kp, Bp, Kr, Br, Kq, Bq) = e.g., Kp=600I3, Bp=40I3, Kr=70I3, Br=5I3, Kq=6I7, Bq=4.5I7 (cocktail); Kp=1600I3, Bp=120I3, Kr=70I3, Br=5I3, Bq=4.5I7…
    Chosen by hand or trial-and-error for each experiment; no systematic selection method is provided (Section 5.4 admits this). The values are free choices needed for the demonstrations to work.
  • Singularity threshold for workspace analysis = singular value of Lambda^{-1} <= 0.03
    The claim that 30% of the workspace is inaccessible depends on this threshold (Figure 5 caption). The threshold is chosen without stated justification and determines the headline percentage.
assumptions (4)
  • domain assumption The environment is a mechanical admittance, enabling linear superposition of mechanical impedances.
    Invoked in Section 3.4.1.2 for the superposition principle. If the environment is not an admittance (e.g., a position source or another impedance), superposition may fail.
  • domain assumption Ideal torque actuators and exact gravity compensation.
    Assumed in Section 3.1 and used in all experiments (MuJoCo with zero gravity; KUKA with built-in gravity compensation). Real actuators have dynamics and compensation is imperfect.
  • ad hoc to paper A sufficiently large symmetric positive-definite joint damping matrix Bq guarantees passivity for time-varying module parameters.
    Stated in Section 3.4.2.2 without a quantitative bound. This is load-bearing for all time-varying virtual trajectory experiments and is the paper's weakest analytical point.
  • standard math Standard stability and passivity results: Lyapunov's direct method, LaSalle's invariance principle, and the skew-symmetry property of (M_dot - 2C).
    Used in the stability proofs of Sections 3.4.2.1 and 3.4.2.2. These are accepted background results.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Modular Robot Control with Motor Primitives." pith.science (2026). https://pith.science/paper/HHSXCJJ4

@misc{pith2026250510694,
  author       = {Pith},
  title        = {Pith review of: Modular Robot Control with Motor Primitives},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HHSXCJJ4}},
  note         = {Machine review of arXiv:2505.10694}
}
read the original abstract

Despite a slow neuromuscular system, humans easily outperform modern robot technology, especially in physical contact tasks. How is this possible? Biological evidence indicates that motor control of biological systems is achieved by a modular organization of motor primitives, which are fundamental building blocks of motor behavior. Inspired by neuro-motor control research, the idea of using simpler building blocks has been successfully used in robotics. Nevertheless, a comprehensive formulation of modularity for robot control remains to be established. In this paper, we introduce a modular framework for robot control using motor primitives. We present two essential requirements to achieve modular robot control: independence of modules and closure of stability. We describe key control modules and demonstrate that a wide range of complex robotic behaviors can be generated from this small set of modules and their combinations. The presented modular control framework demonstrates several beneficial properties for robot control, including task-space control without solving Inverse Kinematics, addressing the problems of kinematic singularity and kinematic redundancy, and preserving passivity for contact and physical interactions. Further advantages include exploiting kinematic singularity to maintain high external load with low torque compensation, as well as controlling the robot beyond its end-effector, extending even to external objects. Both simulation and actual robot experiments are presented to validate the effectiveness of our modular framework. We conclude that modularity may be an effective constructive framework for achieving robotic behaviors comparable to human-level performance.

Figures

Figures reproduced from arXiv: 2505.10694 by the authors.

Figure 1
Figure 1. (A) The three Elementary Dynamic Actions (EDA). Submovements (orange box) and oscillations (blue box) correspond to kinematic primitives and mechanical impedances (green box) manage physical interaction. (B) Elements of EDA combined using a Norton equivalent network model. The virtual trajectory x0(t) (yellow box) consists of submovements (orange box) and/or oscillations (blue box), and mechanical impedances Z (gree… view at source ↗
Figure 2
Figure 2. (A) A definition of a module, which consist of a pair of mechanical impedance Z and the virtual trajectory x0 to which the impedance is connected (Nah et al. 2024a). For the virtual trajectory, a combination of discrete and/or rhythmic movements is used. (B) The major modules used for robot control: a module for joint-space (Zq, q0) (Section 3.3.1), a module for task-space position (Zp, p0) (Section 3.3.2), a module… view at source ↗
Figure 3
Figure 3. A two degrees-of-freedom planar robotic manipulator controlled using Equation (17), and its configuration manifold which is a T 2 torus. (A–C) (respectively (D-F)) The robot passing through kinematic singularity (i.e., the straight-arm posture) to change from right-hand (respectively left-hand) to left-hand (respectively right-hand) configuration. In the planar robot diagram, red markers depict p0 for task-space pos… view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: KUKA iiwa14 robotic manipulator controlled using Equation (18). (A,B) Using the three control modules, the whole robot’s workspace can be utilized. For the experiment, iiwa14 singularity1 KUKA application was used. (C,D) Using the three control modules, the robot can s…
Figure 5
Figure 5. Figure 5: Analysis and quantification of kinematic singularity of the KUKA LBR iiwa14 robotic manipulator. (A) The robot’s workspace regions that were inaccessible using conventional methods (Khatib 1987; Chiaverini 1997). Dots depict workspace locations where the singular value…
Figure 6
Figure 6. Figure 6: KUKA iiwa14 robotic manipulator controlled using Equation (18), with the task of maintaining the end-effector’s orientation while following a circular trajectory in task-space. (A,B,C) Result using a zero joint-stiffness matrix Kq = 0. (D,E,F) Result using a symmetric …
Figure 7
Figure 7. Figure 7: (A) The KUKA LBR iiwa14 maintains an external wrench exerted by a 31 kg heavy-load bookshelf. (B) Measured force (top) and torque (bottom) from the force/torque sensor. An ATI Industrial Automation Gamma force/torque sensor was used. (C) Torque as a percentage of satur…
Figure 8
Figure 8. Figure 8: KUKA iiwa14 robotic manipulator shaking a cocktail. The robot was controlled using Equation (18), and Imitation Learning was used to learn the virtual task-space position p0(t) (Appendix C.4.2) and orientation R0(t) (or ⃗q0(t)) (Appendix C.4.3). (A) Robot movement gene…
Figure 9
Figure 9. Figure 9: The KUKA iiwa14 robotic manipulator performing a pouring motion. Robot movements resulting from defining a fixed virtual trajectory p0 and the task-space position p are illustrated for two cases: (A) defined at the robot’s end-effector and (B) defined at the tip of the…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

  1. Adaptive Modular Geometric Control of Robotic Manipulators

    eess.SY 2026-03 unverdicted novelty 7.0 of 10

    An adaptive modular geometric controller for robotic manipulators achieves exponential stability in nominal cases and reduces RMS position error by at least 12.2% versus state-of-the-art methods in simulations while u...

Reference graph

Works this paper leans on

72 extracted references · 65 canonical work pages · cited by 1 Pith paper

  1. [1]

    (2023) Agile catching with whole-body mpc and blackbox policy learning

    Abeyruwan S, Bewley A, Boffi NM, Choromanski KM, D’Ambrosio DB, Jain D, Sanketi PR, Shankar A, Sindhwani V , Singh S et al. (2023) Agile catching with whole-body mpc and blackbox policy learning. In: Learning for Dynamics and Control Conference. PMLR, pp. 851–863. Abu-Dakka FJ and Kyrki V (2020) Geometry-aware dynamic movement primitives. In: 2020 IEEE In...

    work page 2023

  2. [2]

    Duan A, Batzianoulis I, Camoriano R, Rosasco L, Pucci D and Billard A (2024) A structured prediction approach for robot imitation learning

    Springer. Duan A, Batzianoulis I, Camoriano R, Rosasco L, Pucci D and Billard A (2024) A structured prediction approach for robot imitation learning. The International Journal of Robotics Research 43(2): 113–133. Prepared using sagej.cls Modular Robot Control with Motor Primitives 31 Duruisseaux V , Duong TP, Leok M and Atanasov N (2023) Lie group forced ...

    work page 2024

  3. [3]

    2523–2528

    IEEE, pp. 2523–2528. Buchli J, Stulp F, Theodorou E and Schaal S (2011) Learning variable impedance control. The International Journal of Prepared using sagej.cls 30 Journal Title XX(X) Robotics Research 30(7): 820–833. Bullo F (2024) Contraction Theory for Dynamical Systems . 1.2 edition. Kindle Direct Publishing. ISBN 979-8836646806. URL https://fbullo....

    work page 2011

  4. [4]

    Betsch P and Steinmann P (2001) Constrained integration of rigid body dynamics

    Athena scientific. Betsch P and Steinmann P (2001) Constrained integration of rigid body dynamics. Computer methods in applied mechanics and engineering 191(3-5): 467–488. Billard A and Kragic D (2019) Trends and challenges in robot manipulation. Science 364(6446): eaat8414. Billard A, Mirrazavi S and Figueroa N (2022)Learning for adaptive and reactive ro...

    arXiv 2001

  5. [5]

    Bellman R (1966) Dynamic programming

    MIT press. Bellman R (1966) Dynamic programming. Science 153(3731): 34–

    work page 1966

  6. [7]

    IEEE Transactions on Robotics 36(2): 348–365

    Lee T, Wensing PM and Park FC (2019) Geometric robot dynamic identification: A convex programming approach. IEEE Transactions on Robotics 36(2): 348–365. Lee Y (2024) Mmp++: Motion manifold primitives with parametric curve models. IEEE Transactions on Robotics . Leonard I (1996) The matrix exponential. SIAM review 38(3): 507–

    work page 2019

  7. [10]

    Saveriano M, Abu-Dakka FJ, Kramberger A and Peternel L (2023) Dynamic movement primitives in robotics: A tutorial survey

    Springer Science & Business Media. Saveriano M, Abu-Dakka FJ, Kramberger A and Peternel L (2023) Dynamic movement primitives in robotics: A tutorial survey. The International Journal of Robotics Research 42(13): 1133–

    work page 2023

  8. [15]

    Neural networks 21(4): 642–

    Ijspeert AJ (2008) Central pattern generators for locomotion control in animals and robots: a review. Neural networks 21(4): 642–

    work page 2008

Show all 72 references
  1. [16]

    The Handbook of brain theory and neural networks : 579–582

    Jordan MI and Jacobs RA (1995) Modular and hierarchical learning systems. The Handbook of brain theory and neural networks : 579–582. Jordan MI and Rumelhart DE (2013) Forward models: Supervised learning with a distal teacher. In:Backpropagation. Psychology Press, pp. 189–236....

    1995

  2. [20]

    1749–1754

    IEEE, pp. 1749–1754. Fasse ED and Hogan N (1996) Control of physical contact and dynamic interaction. In: Robotics Research: The Seventh International Symposium. Springer, pp. 28–38. Featherstone R (2014) Rigid body dynamics algorithms. Springer. Feix T, Romero J, Schmiedmayer...

    1996 arXiv

  3. [23]

    In: 2021 20th International Conference on Advanced Robotics (ICAR)

    Anand AS, Østvik A, Grøtli EI, Vagia M and Gravdahl JT (2021) Real-time temporal adaptation of dynamic movement primitives for moving targets. In: 2021 20th International Conference on Advanced Robotics (ICAR). IEEE, pp. 261–268. Andreu MG, Ploeger K and Peters J (2024) Beyond...

    2021

  4. [25]

    IEEE Transactions on Robotics 32(5): 1298–1305

    Kronander K and Billard A (2016) Stability considerations for variable impedance control. IEEE Transactions on Robotics 32(5): 1298–1305. Kuindersma S, Deits R, Fallon M, Valenzuela A, Dai H, Permenter F, Koolen T, Marion P and Tedrake R (2016) Optimization- based locomotion p...

    2016 arXiv

  5. [26]

    In: Humanoids 2008-8th IEEE- RAS International Conference on Humanoid Robots

    Park DH, Hoffmann H, Pastor P and Schaal S (2008) Movement reproduction and obstacle avoidance with dynamic movement primitives and potential fields. In: Humanoids 2008-8th IEEE- RAS International Conference on Humanoid Robots. IEEE, pp. 91–98. Park F and Ravani B (1995) Bezie...

    2008

  6. [30]

    Journal of experimental psychology: Human perception and performance 11(6):

    Viviani P and Cenzato M (1985) Segmentation and coupling in complex movements. Journal of experimental psychology: Human perception and performance 11(6):

    1985

  7. [31]

    Karaman S and Frazzoli E (2011) Sampling-based algorithms for optimal motion planning

    McGraw-hill New York. Karaman S and Frazzoli E (2011) Sampling-based algorithms for optimal motion planning. The international journal of robotics research 30(7): 846–894. Kargo WJ and Giszter SF (2000) Rapid correction of aimed movements by summation of force-field primitives...

    2011

  8. [32]

    Trends in Cognitive Sciences

    Grossberg S (1998) The complementary brain: A unifying view of brain specialization and modularity. Trends in Cognitive Sciences . Guiard Y (1993) On fitts’s and hooke’s laws: Simple harmonic movement in upper-limb cyclical aiming. Acta psychologica 82(1-3): 139–159. Guiard Y ...

    1998

  9. [37]

    Bernstein N (1967) The co-ordination and regulation of movements. pergamo. Press, London . Bernstein NA (1935) The problem of interrelation between coordination and localization. Arch Biol Sci 38: 1–35. Berret B, Chiovetto E, Nori F and Pozzo T (2011) Evidence for composite co...

    1967

  10. [41]

    American Mathematical Society, pp. 49–89. Baldwin CY and Clark KB (1999) Design Rules: The Power of Modularity Volume

    1999

  11. [42]

    Nature neuroscience 6(3): 300–308

    d’Avella A, Saltiel P and Bizzi E (2003) Combinations of muscle synergies in the construction of a natural motor behavior. Nature neuroscience 6(3): 300–308. Davidson PR and Wolpert DM (2004) Scaling down motor memories: de-adaptation after motor learning. Neuroscience letters...

    2003

  12. [47]

    In: 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

    Ploeger K and Peters J (2022) Controlling the cascade: Kinematic planning for n-ball toss juggling. In: 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, pp. 1139–1144. Polyakov F (2017) Affine differential geometry and smoothness maximizat...

    2022 arXiv

  13. [49]

    Bullo F and Murray RM (1995) Proportional derivative (PD) control on the Euclidean group

    Springer. Bullo F and Murray RM (1995) Proportional derivative (PD) control on the Euclidean group . PhD Thesis, California Institute of Technology. Burdet E, Osu R, Franklin DW, Milner TE and Kawato M (2001) The central nervous system stabilizes unstable dynamics by learning ...

    1995

  14. [50]

    1585–1590

    IEEE, pp. 1585–1590. Righetti L and Ijspeert AJ (2008) Pattern generators with sensory feedback for the control of quadruped locomotion. In: 2008 IEEE International Conference on Robotics and Automation . IEEE, pp. 819–824. Robinson AC (1958) On the use of quaternions in simul...

    2008

  15. [51]

    Journal of neuroscience 18(23): 10105– 10115

    Santello M, Flanders M and Soechting JF (1998) Postural hand synergies for tool use. Journal of neuroscience 18(23): 10105– 10115. Sastry S (2013) Nonlinear systems: analysis, stability, and control, volume

    1998

  16. [59]

    Frontiers in computational neuroscience 7:

    Sternad D, Marino H, Charles SK, Duarte M, Dipietro L and Hogan N (2013) Transitions between discrete and rhythmic primitives in a unimanual task. Frontiers in computational neuroscience 7:

    2013

  17. [61]

    Acta psychologica 54(1-3): 115–130

    Lacquaniti F, Terzuolo C and Viviani P (1983) The law relating the kinematic and figural aspects of drawing movements. Acta psychologica 54(1-3): 115–130. LaSalle J (1960) Some extensions of liapunov’s second method. IRE Transactions on circuit theory 7(4): 520–527. Latash ML ...

    1983

  18. [62]

    PhD Thesis, Massachusetts Institute of Technology

    Newman WS (1987) High-speed robot control in complex environments. PhD Thesis, Massachusetts Institute of Technology. Nguyen-Tuong D and Peters J (2010) Using model knowledge for learning inverse dynamics. In: 2010 IEEE international conference on robotics and automation. IEEE...

    1987

  19. [63]

    Takegaki M (1981) A new feedback method for dynamic control of manipulators. Trans. ASME, Ser. G, J. Dynamic Systems, Measurement, and Control 103(2): 119–125. Tassa Y (2011) Theory and implementation of biomimetic motor controllers. Citeseer. Theodorou E, Buchli J and Schaal ...

    1981

  20. [71]

    The International Journal of Robotics Research 43(3): 330–353

    Holladay R, Lozano-P ´erez T and Rodriguez A (2024) Robust planning for multi-stage forceful manipulation. The International Journal of Robotics Research 43(3): 330–353. Huang L (2009) Velocity planning for a mobile robot to track a moving target—a potential field approach. Ro...

    2024

  21. [72]

    IEEE Transactions on Systems, Man, and Cybernetics 16(1): 93–101

    Wampler CW (1986) Manipulator inverse kinematic solutions based on vector formulations and damped least-squares methods. IEEE Transactions on Systems, Man, and Cybernetics 16(1): 93–101. Wensing PM, Kim S and Slotine JJE (2017) Linear matrix inequalities for physically consist...

    1986

  22. [90]

    Frontiers in sports and active living 2: 596063

    Stetter BJ, Herzog M, M ¨ohler F, Sell S and Stein T (2020) Modularity in motor control: similarities in kinematic synergies across varying locomotion tasks. Frontiers in sports and active living 2: 596063. Stramigioli S (2001) Modeling and IPC control of interactive mechanica...

    2020

  23. [126]

    Journal of neuroscience 13(2): 467–491

    Giszter SF, Mussa-Ivaldi FA and Bizzi E (1993) Convergent force fields organized in the frog’s spinal cord. Journal of neuroscience 13(2): 467–491. Gomez-Gonzalez S, Neumann G, Sch ¨olkopf B and Peters J (2016) Using probabilistic movement primitives for striking movements. In...

    1993

  24. [127]

    In: Mathematical Theory of Networks and Systems: Proceedings of the MTNS-83 International Symposium Beer Sheva, Israel, June 20–24,

    Brockett RW (2005) Robotic manipulators and the product of exponentials formula. In: Mathematical Theory of Networks and Systems: Proceedings of the MTNS-83 International Symposium Beer Sheva, Israel, June 20–24,

    2005

  25. [131]

    (1991) Potential field methods and their inherent limitations for mobile robot navigation

    Koren Y , Borenstein J et al. (1991) Potential field methods and their inherent limitations for mobile robot navigation. In: Icra. pp. 1398–1404. Koutras L and Doulgeri Z (2020a) A correct formulation for the orientation dynamic movement primitives for robot control in the car...

    1991

  26. [167]

    In: 2021 60th IEEE Conference on Decision and Control (CDC)

    Tsukamoto H, Chung SJ and Slotine JJ (2021) Learning-based adaptive control using contraction theory. In: 2021 60th IEEE Conference on Decision and Control (CDC). IEEE, pp. 2533–

    2021

  27. [191]

    National Aeronautics and Space Administration, Ames Research Center

    Meyer G (1971) Design and global analysis of spacecraft attitude control systems . National Aeronautics and Space Administration, Ames Research Center. Miall RC, Weir DJ, Wolpert DM and Stein J (1993) Is the cerebellum a smith predictor?Journal of motor behavior25(3): 203–216....

    1971

  28. [200]

    Silver D, Lever G, Heess N, Degris T, Wierstra D and Riedmiller M (2014) Deterministic policy gradient algorithms

    Springer. Silver D, Lever G, Heess N, Degris T, Wierstra D and Riedmiller M (2014) Deterministic policy gradient algorithms. In: International conference on machine learning. Pmlr, pp. 387–

    2014

  29. [228]

    Mechanical, Electrical and Electromechanical Applica- tions

    Ortega R and Nicklasson PJ (2013) Passivity-based control of euler- lagrange systems: Mechanical, electrical and electromechani- cal. Mechanical, Electrical and Electromechanical Applica- tions . Ortega R, Van Der Schaft A, Castanos F and Astolfi A (2008) Control by interconne...

    2013

  30. [269]

    In: 2012 IEEE/RSJ international conference on intelligent robots and systems

    Todorov E, Erez T and Tassa Y (2012) Mujoco: A physics engine for model-based control. In: 2012 IEEE/RSJ international conference on intelligent robots and systems. IEEE, pp. 5026–

    2012

  31. [274]

    Journal of cognitive Neuroscience 3(3): 220–230

    Flash T and Henis E (1991) Arm trajectory modifications during reaching towards visual targets. Journal of cognitive Neuroscience 3(3): 220–230. Flash T and Hochner B (2005) Motor primitives in vertebrates and invertebrates. Current opinion in neurobiology15(6): 660–666. Flash...

    1991

  32. [279]

    In: International Design Engineering Technical Conferences and Computers and Information in Engineering Conference , volume 51814

    Mu˜noz Osorio JD, Fiore MD and Allmendinger F (2018) Operational space formulation under joint constraints. In: International Design Engineering Technical Conferences and Computers and Information in Engineering Conference , volume 51814. American Society of Mechanical Enginee...

    2018

  33. [289]

    In: Proceedings 2001 ICRA

    Bruyninckx H (2001) Open robot control software: the orocos project. In: Proceedings 2001 ICRA. IEEE international conference on robotics and automation (Cat. No. 01CH37164), volume

    2001

  34. [295]

    In: Proceedings 2006 IEEE International Conference on Robotics and Automation,

    Park J and Khatib O (2006) Contact consistent control framework for humanoid robots. In: Proceedings 2006 IEEE International Conference on Robotics and Automation,

    2006

  35. [330]

    Physica D: Nonlinear Phenomena 216(2): 269–281

    Righetti L, Buchli J and Ijspeert AJ (2006) Dynamic hebbian learning in adaptive frequency oscillators. Physica D: Nonlinear Phenomena 216(2): 269–281. Righetti L and Ijspeert AJ (2006) Programmable central pattern generators: an application to biped locomotion control. In: Pr...

    2006

  36. [342]

    In: 2012 12th IEEE-RAS international conference on humanoid robots (humanoids

    Ernesti J, Righetti L, Do M, Asfour T and Schaal S (2012) Encoding of periodic and their transient motions by a single dynamic movement primitive. In: 2012 12th IEEE-RAS international conference on humanoid robots (humanoids

    2012

  37. [381]

    Flash T (1987) The control of hand equilibrium trajectories in multi-joint arm movements.Biological cybernetics 57(4): 257–

    1987

  38. [395]

    muscleless motor synergies and actions without movements: From motor neuroscience to cognitive robotics

    Simon HA (2012) The architecture of complexity. In: The Roots of Logistics. Springer, pp. 335–361. Sloth C, Kramberger A and Iturrate I (2020) Towards easy setup of robotic assembly tasks. Advanced Robotics 34(7-8): 499–513. Slotine JJ (2006) Modularity, synchronization, and w...

    2012

  39. [398]

    Trends in neurosciences10(4): 170–174

    Hogan N and Flash T (1987) Moving gracefully: quantitative theories of motor coordination. Trends in neurosciences10(4): 170–174. Hogan N, Krebs HI, Rohrer B, Palazzolo JJ, Dipietro L, Fasoli SE, Stein J, Hughes R, Frontera WR, Lynch D et al. (2006) Motions or muscles? some be...

    1987

  40. [472]

    IEEE Transactions on Automatic Control 44(10): 1943–1946

    Caccavale F, Siciliano B and Villani L (1999b) Robot impedance control with nondiagonal stiffness. IEEE Transactions on Automatic Control 44(10): 1943–1946. Calinon S (2020a) Gaussians on riemannian manifolds: Applica- tions for robot learning and adaptive control. IEEE Roboti...

    1999 arXiv

  41. [512]

    In: 2024 IEEE International Conference on Robotics and Automation (ICRA)

    Li Y , Zhang Y , Razmjoo A and Calinon S (2024) Representing robot geometry as distance fields: Applications to whole-body manipulation. In: 2024 IEEE International Conference on Robotics and Automation (ICRA). IEEE, pp. 15351–15357. Lillicrap TP, Hunt JJ, Pritzel A, Heess N, ...

    2024 arXiv

  42. [653]

    Neural computation 25(2): 328–373

    Ijspeert AJ, Nakanishi J, Hoffmann H, Pastor P and Schaal S (2013) Dynamical movement primitives: learning attractor models for motor behaviors. Neural computation 25(2): 328–373. Ijspeert AJ, Nakanishi J and Schaal S (2002b) Learning rhythmic movements by demonstration using ...

    2013

  43. [742]

    Nature neuroscience 7(9): 907–915

    Todorov E (2004) Optimality principles in sensorimotor control. Nature neuroscience 7(9): 907–915. Todorov E (2005) Stochastic optimal control and estimation meth- ods adapted to the noise characteristics of the sensorimotor system. Neural computation 17(5): 1084–1108. Todorov...

    2004

  44. [828]

    Journal of Experimental Psychology: Human Perception and Performance 21(1):

    Viviani P and Flash T (1995) Minimum-jerk, two-thirds power law, and isochrony: converging approaches to movement planning. Journal of Experimental Psychology: Human Perception and Performance 21(1):

    1995

  45. [839]

    IEEE Transactions on Robotics and Automation 13(4): 546–556

    Fasse ED and Broenink JF (1997) A spatial impedance controller for robotic manipulation. IEEE Transactions on Robotics and Automation 13(4): 546–556. Fasse ED and Gosselin C (1998) On the spatial impedance control of gough-stewart platforms. In: Proceedings. 1998 IEEE Internat...

    1997

  46. [1143]

    BioEssays 22(11): 1043–1045

    Schlosser G and Thieffry D (2000) Modularity in development and evolution. BioEssays 22(11): 1043–1045. Schmidt M and Lipson H (2009) Distilling free-form natural laws from experimental data. science 324(5923): 81–85. Schulman J, Levine S, Abbeel P, Jordan M and Moritz P (2015...

    2000 arXiv

  47. [1184]

    In: 2019 International Conference on Robotics and Automation (ICRA)

    Saveriano M, Franzel F and Lee D (2019) Merging position and orientation motion primitives. In: 2019 International Conference on Robotics and Automation (ICRA) . IEEE, pp. 7041–7047. Schaal S (1999) Is imitation learning the route to humanoid robots? Trends in cognitive scienc...

    2019

  48. [1400]

    Annals of the New York Academy of Sciences 860(1): 360–376

    Dimitrijevic MR, Gerasimenko Y and Pinter MM (1998) Evidence for a spinal central pattern generator in humans a. Annals of the New York Academy of Sciences 860(1): 360–376. Do Carmo MP and Flaherty Francis J (1992)Riemannian geometry, volume

    1998

  49. [1460]

    Shaw S, Abbatematteo B and Konidaris G (2022) Rmps for safe impedance control in contact-rich manipulation

    Shadmehr R and Krakauer JW (2008) A computational neu- roanatomy for motor control.Experimental brain research185: 359–381. Shaw S, Abbatematteo B and Konidaris G (2022) Rmps for safe impedance control in contact-rich manipulation. In: 2022 International Conference on Robotics...

    2008

  50. [1477]

    (2000) Principles of neural science , volume

    Kandel ER, Schwartz JH, Jessell TM, Siegelbaum S, Hudspeth AJ, Mack S et al. (2000) Principles of neural science , volume

    2000

  51. [1882]

    The Equivalence of Second-Order Impedance Control and Proportional Gain Explicit Force Control

    Wolpert DM and Kawato M (1998) Multiple paired forward and inverse models for motor control. Neural networks 11(7-8): 1317–1329. Wolpert DM, Miall RC and Kawato M (1998) Internal models in the cerebellum. Trends in cognitive sciences 2(9): 338–347. Won J and Hogan N (1995) Sta...

    1998 arXiv

  52. [1911]

    CRC press

    Strogatz SH (2018) Nonlinear dynamics and chaos: with applications to physics, biology, chemistry, and engineering . CRC press. Stulp F, Raiola G, Hoarau A, Ivaldi S and Sigaud O (2013) Learning compact parameterized skills with a single regression. In:2013 13th IEEE-RAS Inter...

    2018

  53. [1983]

    Springer, pp. 120–129. Brown TG (1911) The intrinsic factors in the act of progression in the mammal. Proceedings of the Royal Society of London. Series B, Containing Papers of a Biological Character 84(572): 308–319. Brown TG (1912) The factors in rhythmic activity of the ner...

    1911

  54. [1991]

    arXiv preprint arXiv:2501.09198

    Nah MC, Lachner J, Hogan N and Slotine JJ (2025) Combining movement primitives with contraction theory. arXiv preprint arXiv:2501.09198 . Nah MC, Lachner J, Tessari F and Hogan N (2024b) On the modularity of elementary dynamic actions. In: 2024 IEEE/RSJ International Conferenc...

    2025 arXiv

  55. [2006]

    1963–1969

    IEEE, pp. 1963–1969. Park SW, Marino H, Charles SK, Sternad D and Hogan N (2017) Moving slowly is hard for humans: limitations of dynamic primitives. Journal of neurophysiology 118(1): 69–83. Pastor P, Hoffmann H, Asfour T and Schaal S (2009) Learning and generalization of mot...

    2017

  56. [2012]

    IEEE, pp. 57–64. Ernst MO and Banks MS (2002) Humans integrate visual and haptic information in a statistically optimal fashion. Nature 415(6870): 429–433. Faraji S and Ijspeert AJ (2017) Singularity-tolerant inverse kinemat- ics for bipedal robots: An efficient use of computa...

    2002

  57. [2015]

    Advances in neural information processing systems

    Alvarez M, Peters J, Lawrence N and Sch¨olkopf B (2010) Switched latent force models for movement segmentation. Advances in neural information processing systems

    2010

  58. [2084]

    Journal of Motor Behavior 28(2): 165–183

    Schaal S, Atkeson CG and Sternad D (1996) One-handed juggling: A dynamical approach to a rhythmic movement task. Journal of Motor Behavior 28(2): 165–183. Schaal S, Mohajerian P and Ijspeert A (2007) Dynamics systems vs. optimal control—a unifying view. Progress in brain resea...

    1996

  59. [2538]

    In: 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

    Tulbure A and Khatib O (2020) Closing the loop: Real- time perception and control for robust collision avoidance with occluded obstacles. In: 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) . IEEE, pp. 5700–5707. Ude A, Nemec B, Petri ´c T and M...

    2020

  60. [4649]

    IEEE transactions on rehabilitation engineering 6(1): 75–87

    Krebs HI, Hogan N, Aisen ML and V olpe BT (1998) Robot- aided neurorehabilitation. IEEE transactions on rehabilitation engineering 6(1): 75–87. Krizhevsky A, Sutskever I and Hinton GE (2012) Imagenet classification with deep convolutional neural networks. Advances in neural in...

    1998

  61. [5033]

    Nature neuroscience 5(11): 1226–1235

    Todorov E and Jordan MI (2002) Optimal feedback control as a theory of motor coordination. Nature neuroscience 5(11): 1226–1235. Tresch MC, Saltiel P and Bizzi E (1999) The construction of movement by the spinal cord. Nature neuroscience 2(2): 162–

    2002

  62. [5159]

    (1999) Reinforcement learning

    Sutton RS, Barto AG et al. (1999) Reinforcement learning. Journal of Cognitive Neuroscience 11(1): 126–134. Tagliabue M and McIntyre J (2014) A modular theory of multisensory integration for motor control. Frontiers in computational neuroscience 8:

    1999

  63. [6050]

    Courier Corporation

    Kirk DE (2004) Optimal control theory: an introduction . Courier Corporation. Kitano H (2004) Biological robustness. Nature Reviews Genetics 5(11): 826–837. Kober J, Bagnell JA and Peters J (2013) Reinforcement learning in robotics: A survey. The International Journal of Robot...

    2004

  64. [8304]

    Motor control 8(4): 472–483

    Rohrer B, Fasoli S, Krebs HI, V olpe B, Frontera WR, Stein J and Hogan N (2004) Submovements grow larger, fewer, and more blended during stroke recovery. Motor control 8(4): 472–483. Ronsse R, Sternad D and Lefevre P (2009) A computational model for rhythmic and discrete movem...

    2004

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.