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 →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
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.
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
- 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.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
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)
- [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.
- [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.
- [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)
- [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.
- [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.
- [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.
- [Reference [10]] The journal name in reference [10] is misspelled as 'Scoeity'; it should be 'Society'.
Circularity Check
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
assumptions (6)
- domain assumption All manifolds, vector fields, and maps are assumed analytic ('smooth' means analytic, Section 1).
- standard math All affine connections are assumed symmetric, i.e., torsion-free (Section 1).
- 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.
- 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.
- 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.
- domain assumption The vector fields in the symmetric-product closure are complete (Section 3, before Theorem 4).
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.
Reference graph
Works this paper leans on
-
[1]
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
work page 1985
-
[2]
H. Nijmeijer and A. J. Van der Schaft: Partial symmetries f or nonlinear systems, Math- ematical Systems Theory 1, 79 (1985)
work page 1985
-
[3]
H. Nijmeijer and A. J. Van der Schaft: Nonlinear Dynamical Control Systems , Springer, New York 1995
work page 1995
-
[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)
work page 2003
-
[5]
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)
work page 2002
-
[6]
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
work page 2004
-
[7]
A. D. Lewis: The category of affine connection control syste ms, Proceedings of the 39th IEEE Conference on Decision and Control , (2000)
work page 2000
-
[8]
Sandra Ricardo and Witold Respondek: When is a control sys tem mechanical, Journal of Geometric Mechanics 3, 265 (2010)
work page 2010
Show all 12 references
-
[9]
Kobayashi and K
S. Kobayashi and K. Nomizu: Foundations of Differential Geometry, volume 1 , Inter- science publishers, New York 1963
1963
-
[10]
H. J. Sussmann: Orbits of families of vector fields and int egrability of distributions, Transactions of the American Mathematical Scoeity 180, 171 (1973)
1973
-
[11]
Yano and S
K. Yano and S. Ishihara: Tangent and Contangent Bundles: Differential Geometry , Marcel Dekker Inc., the university of Michigan 1973
1973
-
[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)
2007
Reviewed August 14, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.