Pith. sign in

REVIEW 3 major objections 4 minor 12 references

Quotients of affine connection control systems

T0 review · 3 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

Pith's one-line read Mechanical quotients exist exactly when an invariant distribution is present

desk verdict The local quotient theorem is a solid new result, but the global theorem is misstated as an iff and needs a completeness hypothesis or a weaker conclusion. read the letter →

arxiv 1908.01912 v1 pith:L4JJKRAL submitted 2019-08-06 math.DG

classification math.DG MSC 53C0593B2770Q0558A30
keywords affineconnectioncontrolsystemsmechanicalquotientgeodesicaccessibilityinvariantdistributionssymmetricproductfiberbundlesgeometric
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 asks when an affine connection control system, a second-order mechanical system driven by forces, can be quotiented to another mechanical control system without losing the mechanical structure. For geodesically accessible systems—those whose control fields generate the full tangent space under the symmetric product—a local quotient exists exactly when the configuration manifold carries an involutive distribution $D$ satisfying three conditions: the affine connection restricts to $D$, the curvature satisfies $R(X,v)v \in \Gamma(D)$ for every $X \in \Gamma(D)$ and $v \in \Gamma(TQ)$, and each control field brackets into $D$. The quotient map must then be the tangent map of a submersion between configuration spaces, and the quotient system is again geodesically accessible. Globally, under a completeness assumption on the vector fields generated from the controls by the symmetric product, the same conditions characterize quotients for which the configuration manifold is a fiber bundle. If the characterization is right, it gives a precise geometric test for when reducing a mechanical control system preserves its mechanical structure.

What carries the argument

The load-bearing object is an invariant distribution $D$ on the configuration manifold, lifted to the tangent bundle as $\tilde D = \operatorname{span}\{D^{vlft}, [S, D^{vlft}]\}$, where $S$ is the geodesic spray and $D^{vlft}$ denotes the vertical lifts of sections of $D$. The three conditions on $D$ make $\tilde D$ invariant under the geodesic spray and under the control fields, and they ensure the quotient inherits an affine connection structure. The proof is carried by the bracket identities $[S, X^{vlft}] = -X \oplus \nabla_{v_q} X$ and $[S,[S,X^{vlft}]] = -2\nabla_{v_q}X \oplus (R(X,v_q)v_q + \mathrm{ver}(\nabla^H_{v_q^H}(\nabla_{v_q}X)^{vlft}))$, together with an orbit theorem argument that shows the quotient map is a submersion.

What would settle it

Exhibit a geodesically accessible affine connection control system on a noncompact manifold whose control vector fields are not complete, yet which still admits a global quotient map satisfying (7). If the configuration manifold is not a fiber bundle over the quotient configuration space, then Theorem 4's global characterization fails as stated; a concrete search would use polynomial control fields on $\mathbb{R}^n$ that blow up in finite time but generate $T\mathbb{R}^n$ under symmetric products.

Watch

Extended reading notes

Core claim

The central result is Theorem 1: question (i) is solvable if and only if there exists an involutive distribution $D$ on the configuration manifold such that the affine connection restricts to $D$, the curvature satisfies $R(X,v)v \in \Gamma(D)$ for all $X \in \Gamma(D)$ and $v \in \Gamma(TQ)$, and $[g_i,D] \subseteq D$ for each control field $g_i$. Whenever such a quotient exists, the quotient map has the form $T\Phi$ for a submersion $\Phi$ between configuration manifolds, and the quotient mechanical control system is itself a geodesically accessible affine connection system. The global version, Theorem 4, states that question (ii) is solvable if and only if the configuration manifold has a fiber bundle structure whose vertical distribution satisfies the same conditions, with the added assumption that the vector fields in the symmetric-product closure are complete.

Load-bearing premise

The global characterization in Theorem 4 rests on the assumption that every vector field in the symmetric-product closure of the control fields is complete, a condition stated just before the theorem rather than in it, and on analyticity throughout; if completeness fails, the fiber-bundle conclusion of the only-if direction need not follow.

Editorial extensions

If this is right

  • A local quotient that preserves mechanical structure exists for a geodesically accessible affine connection system precisely when the three distribution conditions hold.
  • Every such quotient map is the tangent lift of a submersion between configuration manifolds, so the structure-preserving quotient is a morphism of the mechanical category.
  • The quotient of a geodesically accessible system is again geodesically accessible, and a fully actuated mechanical system quotients to a fully actuated system.
  • On an irreducible Riemannian manifold with its Levi-Civita connection, no nontrivial local mechanical quotient exists, because the only totally geodesic distribution is the entire tangent space.
  • Globally, the configuration manifold must be a fiber bundle over the quotient configuration manifold, with its vertical distribution satisfying the same conditions, provided the generated vector fields are complete.

Reading between the lines

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

  • If the completeness assumption in Theorem 4 fails, the local result should persist but the fiber-bundle conclusion likely weakens to a foliation or an immersion statement, since the only-if proof invokes a theorem requiring completeness.
  • The three distribution conditions could be used as a practical reduction test in geometric mechanics: before quotienting an underactuated system, look for involutive distributions closed under the connection and curvature, not merely control-invariant subspaces.
  • The worked non-geodesically-accessible example suggests that without geodesic accessibility, the quotient can acquire drift and dissipation-like terms, so the geodesically accessible class is the natural setting for structure-preserving reduction.
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

3 major / 4 minor

Summary. The paper studies quotients of affine connection control systems (ACCSs) that are themselves mechanical control systems. It distinguishes a local question (i) and a global question (ii): when does a geodesically accessible ACCS admit, locally or globally, a quotient mechanical control system. The main local result, Theorem 1, characterizes solvability of (i) by the existence of an involutive distribution D on the configuration manifold such that the affine connection restricts to D, the curvature condition R(X,v)v lies in D for all X in D and v in TQ, and the control vector fields satisfy [g_i,D] subset D. The global result, Theorem 4, adds that Q is a fiber bundle whose vertical distribution satisfies the same conditions. The paper also proves structural corollaries: quotient maps must be tangent maps of submersions, quotient systems inherit geodesic accessibility, and it gives an example showing that non-geodesically accessible systems need not have mechanical quotients that are affine connection systems.

Significance. The problem is natural and the local characterization is potentially useful: it provides an intrinsic distributional test for the existence of mechanical quotients and identifies the resulting quotient map as the tangent map of a configuration-space submersion. The paper builds on established external results (Sussmann's orbit theorem, the Ricardo-Respondek characterization, and McKay's fiber-bundle theorem) rather than introducing ad-hoc assumptions, and the example in Section 2 correctly marks the boundary of the local theorem. However, the global theorem as stated is not correct without an explicit completeness hypothesis, and several central proof steps are asserted rather than demonstrated. The results are therefore promising but not publishable in their current form; the difficulties appear fixable within the manuscript's scope.

major comments (3)
  1. [Section 3, Theorem 4 and the preceding paragraph] Theorem 4 is stated as an unqualified 'if and only if', but its only-if proof invokes Theorem 3, whose hypotheses include completeness of the vector fields in both families. That completeness assumption is introduced only in the preceding paragraph and is not repeated in the theorem statement. Without it the theorem is false. For example, take Q = R^2 \ {(0,0)} with the flat connection and controls g1 = ∂x, g2 = ∂y. The system is geodesically accessible, and τ(x,y,v_x,v_y) = (x,v_x) is a global quotient map onto the flat affine connection control system on R, so question (ii) is solvable. Yet the induced configuration map Φ(x,y) = x is not a fiber bundle: the fiber over 0 is R \ {0}, whereas nearby fibers are R. The vertical distribution D = span{∂y} is also not the vertical distribution of any fiber bundle. Thus the theorem must either include the completeness assumption in its statement or explicitly state that it is a standing assumption for the whole section, and it should also specify completeness of the vector fields of the quotient system.
  2. [Section 2, Proposition 1 proof] The proof asserts that, because the system is geodesically accessible, the orbit of {S, g1^vlft, ..., gm^vlft} equals the neighborhood U. Sussmann's orbit theorem gives an immersed submanifold, and openness of the orbit is an additional statement that needs a rank argument or a precise citation. The next step, that the orbit of the quotient system equals V because τ is surjective, also needs a sentence explaining how τ(U) = V and the invariance of orbits under τ produce the claimed equality in V. These points are load-bearing because Proposition 1 is the bridge between question (i) and the distributional conditions used in Theorem 1.
  3. [Section 2, proof of Theorem 1] In the sufficiency part, the proof states that the distribution ~D = span{D^vlft, [S,D^vlft]} is involutive 'by computing' and then asserts that in adapted coordinates the quotient system has the form (28), with Christoffel symbols and control components depending only on the surviving coordinates (x^{k+1}, ..., x^n). This is a central step: the independence properties are exactly what make the quotient an affine connection control system on the reduced configuration space. The derivation should be written out, or a lemma given, showing that conditions (24), (26), and (27) imply the required coordinate independence. As written, the reader cannot verify the sufficiency claim directly.
minor comments (4)
  1. [Throughout] Several theorem-like environments have duplicated titles, e.g. 'Theorem 1 Theorem Question (i) ...' and 'Corollary 1 Corollary ...' in the text; these are formatting errors that should be cleaned up.
  2. [Introduction and Section 3] The blanket analyticity assumption is stated in the introduction and is used implicitly in Proposition 1; it would help to repeat this hypothesis in the statements of Theorem 1 and Theorem 4 so the reader does not have to infer it from the prose.
  3. [Section 2, proof of Proposition 3] The notation in the coordinates (V,z^i) and the statement span{V1,...,Vn} = span{∂/∂y1,...,∂/∂yn} is confusing: the V_i are vector fields on TQ, and the y^i are base coordinates, so the identification should be explained explicitly.
  4. [Reference [10]] The journal name in reference [10] is misspelled as 'Scoeity'; it should be 'Society'.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular reasoning found; the derivation uses external theorems and parameter-free arguments.

full rationale

The paper's central claims (Theorem 1 and Theorem 4) are derived from previously established external results: Sussmann's orbit theorem (Theorem [10]), the Ricardo-Respondek characterization of mechanical systems (Theorem [8]), and Mckay's fiber bundle theorem (Theorem [12]). There are no fitted parameters, no quantities are renamed as predictions, and no load-bearing step is defined in terms of the conclusion it is supposed to establish. The equivalence proofs in Theorem 1 proceed in the standard way: the 'only if' direction derives the distribution D from an invariant distribution on TQ using bracket computations and equations (10)-(12), and the 'if' direction constructs an invariant distribution from D. Neither direction assumes the conclusion. Theorem 4 invokes Mckay's theorem, which is an external result, to conclude the fiber bundle structure; the completeness assumption needed by Mckay's theorem is mentioned in the paragraph before Theorem 4. Whether that assumption is adequately incorporated into the theorem statement is a correctness and rigor concern, not a circularity concern. The paper contains no self-citations by the authors and no appeal to an author-imported uniqueness theorem. The suspected weakness concerning the completeness assumption affects the validity of the global 'if and only if' in some cases, but it does not make the derivation circular.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

The paper introduces no new physical entities or fitted constants. All assumptions are either standard differential geometry background, the explicitly stated analyticity and symmetry conventions, or the completeness condition for the global theorem. The most fragile assumption is the completeness condition, because it is load-bearing for Theorem 4 and absent from the theorem statement.

assumptions (6)
  • domain assumption All manifolds, vector fields, and maps are assumed analytic ('smooth' means analytic, Section 1).
    Needed for Sussmann's Orbit Theorem and the fiber bundle map theorem to yield open orbits and submersions; the paper notes some results would hold in C-infinity but not all.
  • standard math All affine connections are assumed symmetric, i.e., torsion-free (Section 1).
    The geodesic spray and the bracket formulas (10)-(12) use the symmetric connection; the quotient mechanical structure is defined for symmetric connections.
  • standard math Sussmann's Orbit Theorem: the tangent space of the orbit of an analytic family of vector fields equals the Lie algebra generated by the family at each point.
    Used in Proposition 1 to show that the quotient map is a submersion when the orbit fills a neighborhood.
  • standard math Ricardo-Respondek Theorem (Theorem 2 in the paper): characterization of when a control system is locally state equivalent to a geodesically accessible mechanical system.
    Used in Proposition 2 and Remark 1 to prove that the quotient inherits geodesic accessibility.
  • standard math Mckay's Theorem (Theorem 3 in the paper): a map pushing forward complete vector fields and mapping orbits to orbits is a fiber bundle map on each orbit.
    Used in Theorem 4 to upgrade the global quotient map to a fiber bundle, requiring completeness of vector fields in the symmetric product closure.
  • domain assumption The vector fields in the symmetric-product closure are complete (Section 3, before Theorem 4).
    Required to apply Mckay's theorem in the only-if part of Theorem 4; this assumption is not stated inside Theorem 4 itself.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Quotients of affine connection control systems." pith.science (2026). https://pith.science/paper/L4JJKRAL

@misc{pith2026190801912,
  author       = {Pith},
  title        = {Pith review of: Quotients of affine connection control systems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/L4JJKRAL}},
  note         = {Machine review of arXiv:1908.01912}
}
read the original abstract

In this paper, we investigate the existence of a subclass of quotients of affine connection control systems, which preserve the mechanical structures. Both local and global sufficient and necessary conditions are given for the geodesically accessible affine connection control systems such that they can admit this subclass of quotients. The structural properties of the quotient map and the quotient mechanical control system are discussed.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

12 extracted references · 12 canonical work pages

  1. [1]

    Grizzle and Steven I

    Jessy W. Grizzle and Steven I. Marcus: The structure of non linear control systems pos- sessing symmetries, IEEE Transactions on Automatic Control 3, 248 (1985). Quotients of affine connection control systems 15

  2. [2]

    Nijmeijer and A

    H. Nijmeijer and A. J. Van der Schaft: Partial symmetries f or nonlinear systems, Math- ematical Systems Theory 1, 79 (1985)

  3. [3]

    Nijmeijer and A

    H. Nijmeijer and A. J. Van der Schaft: Nonlinear Dynamical Control Systems , Springer, New York 1995

  4. [4]

    Pappas: Abstractions of Hamil tonian control systems, Automatica 39, 2025 (2003)

    Paulo Tabuada and George J. Pappas: Abstractions of Hamil tonian control systems, Automatica 39, 2025 (2003)

  5. [5]

    Pappas and Pedro Lima: Composing abstractions of hybrid systems, in Hybrid systems: Computation and Control, Claire J

    Paulo Tabuada, George J. Pappas and Pedro Lima: Composing abstractions of hybrid systems, in Hybrid systems: Computation and Control, Claire J. Toml in and Mark R. Greenstreet eds., Lecture Notes in Computer Science 2289, 436 (2002)

  6. [6]

    Bullo and A

    F. Bullo and A. D. Lewis: Geometric Control of Mechanical Systems. Modeling, Analys is and Design for Simple Mechanical Control Systems , Springer, New York 2004

  7. [7]

    A. D. Lewis: The category of affine connection control syste ms, Proceedings of the 39th IEEE Conference on Decision and Control , (2000)

  8. [8]

    Sandra Ricardo and Witold Respondek: When is a control sys tem mechanical, Journal of Geometric Mechanics 3, 265 (2010)

Show all 12 references
  1. [9]

    Kobayashi and K

    S. Kobayashi and K. Nomizu: Foundations of Differential Geometry, volume 1 , Inter- science publishers, New York 1963

  2. [10]

    H. J. Sussmann: Orbits of families of vector fields and int egrability of distributions, Transactions of the American Mathematical Scoeity 180, 171 (1973)

  3. [11]

    Yano and S

    K. Yano and S. Ishihara: Tangent and Contangent Bundles: Differential Geometry , Marcel Dekker Inc., the university of Michigan 1973

  4. [12]

    Mckay: Sussmann’s orbit theorem and maps, Differential Geometry and its Applica- tions 25, 277 (2007)

    B. Mckay: Sussmann’s orbit theorem and maps, Differential Geometry and its Applica- tions 25, 277 (2007)

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.