Pith. sign in

REVIEW 3 major objections 5 minor 57 references

The Dodecacopter: a Versatile Multirotor System of Dodecahedron-Shaped Modules

T0 review · 3 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read A dodecahedron-shaped module can assemble into flat arrays, stiff tetrahedra, and fully actuated six-degree-of-freedom aircraft; up to sixteen modules have flown.

desk verdict Genuinely new modular rotorcraft geometry with working prototypes; the geometric core is solid, but the convexity appendix is wrong and the efficiency data cannot support the 3D scaling claim. read the letter →

arxiv 2504.16475 v1 pith:WOZBDU7Y submitted 2025-04-23 cs.RO

Kévin Garanger , Thanakorn Khamvilai , Jeremy Epps , Eric Feron This is my paper · ORCID
classification cs.RO
keywords modularreconfigurableUAVdodecahedronmodulefullyactuatedmultirotorcontrolallocationmixed-integerprogrammingspaceframestiffnessrotorwakeinteractionprototypeflighttests
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

The paper is trying to establish that a single module—a regular dodecahedron frame with a central fixed-pitch propeller—can serve as the building block for a wide range of multirotor vehicles. Its connection method lets modules join at selected vertices, and the paper proves that every reachable assembly sits on a discrete three-dimensional lattice, which turns the search for good vehicle shapes into a mixed-integer optimization problem. The central payoff is that identical hardware can form ordinary flat quadrotors and hexarotors, stiff layered tetrahedral structures, and fully actuated six-degree-of-freedom aircraft able to push thrust in any direction. The paper also reports hover flights of up to sixteen modules in six different configurations, including a four-layer tetrahedral vehicle.

What carries the argument

The load-bearing mechanism is the restricted connection: two modules are joined by choosing, in each module, a pair of non-adjacent vertices that lie on a common face, and aligning those pairs so the two faces are coplanar. When the eight vertices used for connections are chosen to form an inscribed cube, the resulting translation vectors generate the three-dimensional lattice $\mathcal{P}$ of Fact II.4, with basis $(1,1,0)$, $(1,0,1)$, $(0,1,1)$. This lattice does three jobs: it guarantees that frames of different modules never intersect, it makes every compatible module position representable by integer coordinates, and it reduces the configuration space to a discrete set when combined with the finite set $H$ of feasible rotor orientations. That discreteness is what turns vehicle design into a mixed-integer program with linear and second-order-conic constraints, and it is why the same module can be optimized into a stiff tetrahedron or a high-yaw-authority 12-rotor 6-DOF vehicle.

What would settle it

Run hover or thrust-stand tests on 1-, 2-, 3-, and 4-layer Dodecacopter configurations with several repeated trials per configuration, keeping rotor overlap geometry fixed; if the average per-module power penalty grows substantially with each added layer beyond the first, the assumption that distant wake interactions are small is false. A more direct check is to measure rotor thrust with a second rotor placed one, two, and three layers away at fixed spacing and test whether the induced loss decays with separation.

Watch

Extended reading notes

Core claim

The central claim is that the dodecahedron's rotational symmetry, combined with a connection rule based on pairs of vertices lying on a common face but not adjacent, gives a modular rotorcraft design whose configuration space is both rich and computationally tractable. Restricted to the eight vertices of an inscribed cube, the connection translations generate a lattice whose basis includes the vectors $(1,1,0)$, $(1,0,1)$, and $(0,1,1)$; any compatible module position lies on this lattice, and any orientation is a rotation from the 60-element symmetry group of the dodecahedron. Because positions are discrete and orientations finite, a configuration can be written with binary variables, and the paper uses this to formulate mixed-integer programs that optimize control authority and structural stiffness. The paper then verifies the concept experimentally: six configurations with 4 to 16 modules were hovered, including a 6-DOF hexarotor and layered tetrahedra, showing that the same hardware physically realizes flat, three-dimensional, and fully actuated vehicles.

Load-bearing premise

The load-bearing premise is that wake losses come mainly from rotors in adjacent layers, so adding farther layers costs little efficiency; the paper's own evidence is a single hover comparison with no repeated trials.

Editorial extensions

If this is right

  • A single kit of identical modules can replace several monolithic aircraft, since the same module recreates quadrotor and hexarotor layouts directly.
  • Three-dimensional assemblies are stiffer than flat arrays with the same module count; for four modules, the tetrahedral quadrotor's worst-case compliance is about a quarter of the flat quadrotor's.
  • Fully actuated hovering is achievable without tilting mechanisms by choosing module orientations from the dodecahedron's rotational symmetries, as in the optimized 12-rotor example.
  • Because overactuated configurations admit many control allocation matrices, the paper's convex programs let a designer trade reachable thrust and torque authority against expected power consumption.
  • The prototype flights of up to sixteen modules, including layered tetrahedral shapes, indicate the concept is practically flyable, not only theoretically valid.

Reading between the lines

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

  • If the hover data generalize, then very tall stacks of Dodecacopter modules may stay close to flat-array efficiency, making large 3D vehicles practical; this is an inference from the paper's one-dataset observation, not a proven result.
  • The lattice formulation is tied to connections through the eight cube vertices; allowing all 3600 possible face-pair connections would likely yield different position sets with different rigidity and actuation properties that the paper does not explore.
  • The control-allocation and stiffness formulations are stated for arbitrary rotor positions, so they should transfer to any overactuated multirotor, modular or not.
  • Because configuration data produce both stiffness and actuation matrices, one could extend the pipeline to compute control gains automatically, removing the manual per-configuration tuning that the paper reports as the main practical bottleneck.
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 / 5 minor

Summary. The paper introduces the Dodecacopter, a modular rotorcraft system in which each module is a regular dodecahedron with a fixed-pitch propeller mounted along an axis through two opposite vertices. The proposed connection rule, based on joining two non-adjacent vertices of a face and restricted to a cube subset of dodecahedron vertices, is shown to generate a three-dimensional lattice of admissible module positions and a finite set of rotor orientations. On this basis the authors derive the thrust and torque matrices of arbitrary configurations, propose convex programs for control allocation (maximizing a reachable actuation set and minimizing average power consumption), define structural stiffness indicators via space-frame analysis, and formulate mixed-integer programs for configuration optimization under connectivity, wake-avoidance, and actuation constraints. The paper also reports a prototype and hover tests of six configurations, including flat quadrotors/hexarotors, tetrahedral three-dimensional assemblies, and a six-DOF hexarotor.

Significance. If the results hold, this is a meaningful step beyond two-dimensional modular flight arrays: the same module can form flat arrays, stiff three-dimensional tetrahedral structures, and fully actuated vehicles, and the lattice description makes configuration optimization amenable to mixed-integer programming. The clean geometric characterizations (Fact II.3 and Fact II.4), the structural performance indicators, and the six flown configurations are genuine strengths. The central geometric derivation appears sound. However, the paper's practical-viability argument relies on a wake-interaction inference that is not statistically supported, and the convexity proof for one of the proposed allocation programs contains an algebraic error. Neither issue appears fatal to the core module concept, but both need to be corrected or substantially qualified before the paper can be accepted.

major comments (3)
  1. [Appendix A; Program III.4] The claimed convexity proof for Eq. (46) is not valid as written. Kummer's transformation applied with a = -3/4, b = 1/2, and z = -x gives f(x) = x^{-3/4} e^{-x} 1F1(5/4,1/2,x), not the expression with e^x in Eq. (50). The subsequent second-derivative expression mixes e^{-x} factors with terms that appear inconsistent with either version of f, and the final positivity assertion is stated after 'grouping terms' without a verifiable intermediate derivation. Since Program III.4 relies on convexity of Eq. (15) through this lemma, the power-allocation contribution is not supported as written. I am not asserting the result is false; I am asking for a corrected proof, a reliable citation, or explicit removal of the convexity claim from the main contribution list.
  2. [VI.C; Table II] The conclusion that wake-induced losses are dominated by adjacent-layer interactions, and that non-adjacent interactions are small, is not established by the presented data. Table II reports one normalized motor-input value per configuration, with no repeat trials, error bars, or significance tests. The three tetrahedron configurations differ simultaneously in module count, number of layers, and rotor-overlap geometry, so the 14% penalty relative to flat configurations and the monotone increase from 0.34 to 0.35 to 0.37 cannot be cleanly attributed to adjacent-layer wake. Non-adjacent-layer interactions are never measured in isolation. Because the practical case for larger multi-layer Dodecacopter vehicles rests on this inference, the claim should either be supported by controlled experiments (for example, varying layer separation at constant module count) or explicitly reduced to a speculative observation.
  3. [V.E, Fact V.2] The claim that M_tt C = H can be linearized with the big-M method is incomplete. Big-M linearization of products of binary variables with a continuous variable C requires a priori bounds on C. The programs in Section V do not impose explicit bounds on C; the norm constraints in Program V.1 bound certain combinations of the rows of C, but not the full matrix C unless additional assumptions are stated. Without a bounded feasible set or an explicit bound, the equivalence asserted in Fact V.2 is not established, which weakens the mixed-integer formulation underlying Program V.1. Please add explicit bounds (for example, derived from actuator limits and a chosen M) or prove that the feasible set is bounded.
minor comments (5)
  1. [Section V.G, Program V.1] The displayed Program V.1 is under-specified: the objective minimizes lambda, but no constraint involves lambda, and the inequality uses s without defining it as a variable or as a fixed parameter. Please state the objective and the role of s explicitly.
  2. [Section II.B, Definition II.5] The concatenation operation in Definition II.5 is described as composition, but translations are applied in the original coordinate frame, which is nonstandard. Calling it a concatenation operation rather than composition would prevent confusion.
  3. [Section III.B.3] There are several small typos: 'Moore-Pensore' in Fact III.2, 'an other method' in Section III.B.3, 'CRFP' in Section VI.A (likely CFRP), and 'an arbitrary a 6DOF configuration' in Section V.G. These should be corrected in a revision.
  4. [Section III.C, Figures 9 and 10] The comparison of allocation matrices would be easier to interpret if Figure 9 included error bars or a statement of the number of simulations, and if Figure 10 had labeled axes and visible scale information for both the per-matrix sets and the global reachable set.
  5. [Section VI.B, Eq. (45)] The displayed control allocation matrix for the tetrahedron quadrotor is said to show that the top motor receives three-times-larger yaw commands, but the matrix is shown up to scale; please state explicitly how the matrix is normalized.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the dodecahedron connection lattice, actuation models, optimization programs, and flight tests are self-contained; self-citations are background only and are not load-bearing.

full rationale

The derivation chain is self-contained. The lattice of reachable module positions is constructed from the dodecahedron vertex coordinates, as in Fact II.4 where the vectors (1,1,0), (1,0,1), and (0,1,1) are shown to generate the three-dimensional lattice P, and Fact II.3 characterizes compatibility from the definition of connections and the group structure of S_D and P rather than assuming it. The actuation model in Fact II.6 is the standard thrust/torque proportionality model, and the control-allocation programs are convex reformulations of that model; no parameter is fitted to later flight data and then renamed a prediction. The structural indicators in Section IV are derived from the textbook space-frame elastic stiffness matrix, not from the paper's own conclusions. The MIP formulation in Section V encodes the connection graph, contiguity constraints, wake-exclusion constraints, and stiffness constraints using standard graph and big-M techniques, so the example configurations are outputs of the stated programs rather than inputs. The only self-citations, [8], [25], and [49], are background references about wake interactions and continuing work; they do not carry the central claims that dodecahedron modules generate three-dimensional and fully actuated configurations. The flight-test section is empirical and explicitly hedged: "Although no definitive conclusion can be made based solely on this observation..." That caveat concerns statistical strength, not circularity, and the paper's central geometric and actuation claims do not depend on that inference. No fitted input is relabeled as a prediction, no uniqueness theorem is imported from the authors' prior work, and no result reduces by definition to its own input.

Assumptions & free parameters 4 free parameters · 9 assumptions · 1 invented entities

The central claim rests on standard rigid-body and rotorcraft modeling, the finite symmetry group of the dodecahedron, and a few domain assumptions about wake behavior and structural rigidity. The only ad hoc modeling choice is the Gaussian distribution of command vectors used for the power-consumption objective. Key numerical parameters (coefficients, covariance, beam properties) are not published, which limits reproduction.

free parameters (4)
  • Thrust and torque coefficients kth and kto
    These coefficients in Fact II.6 convert squared rotor speeds to forces and moments. The numerical comparison of allocation matrices in Section III.C requires their values, but none are given, so the reported control-authority and power plots cannot be reproduced as-is.
  • Hover control vector uh
    Program III.1 fixes a hover control vector uh but only suggests examples such as M_tt^+ th; the actual choice changes the reachable-set maximization and is not specified for the comparisons.
  • Covariance Sigma of command vector distribution
    The average power consumption model in Section III.B.3 assumes t ~ N(th, Sigma); Sigma is not specified, so Program III.4's solution depends on an unstated modeling choice.
  • Structural member properties (beam cross-section, material)
    The space-frame stiffness matrices in Section IV require beam geometry and elastic modulus. The reported sigma1 and sigma2 values are normalized to the quadrotor, so absolute properties may cancel, but actual load predictions require them.
assumptions (9)
  • standard math Finite subgroups of SO(3) are exactly cyclic, dihedral, tetrahedral, octahedral, or icosahedral.
    Used in Section II to justify choosing the dodecahedron for maximal rotational symmetry; cited to [26].
  • standard math The regular dodecahedron has rotational symmetry group of order 60.
    Fact II.1, cited to Coxeter [27].
  • domain assumption Each rotor produces thrust and counter-torque proportional to the square of its rotation speed, with constant coefficients.
    Fact II.6 and equations (7)-(8); standard rotorcraft model but ignores wake, inflow, and RPM-dependent effects that the paper's own experiments show matter at the roughly 14 percent level.
  • domain assumption Power consumption of a rotor is proportional to u_i^{3/2}.
    Section III.B.3, justified by momentum theory [34], [35]; used to define the power-minimizing allocation program.
  • ad hoc to paper The high-level actuation commands t are multivariate normal with covariance Sigma.
    Section III.B.3, equation (12); this distributional assumption is introduced specifically to make average power consumption a closed-form function of the allocation matrix.
  • domain assumption Frame joints are perfectly rigid and displacements are small.
    Section IV, standard space-frame analysis assumption [40]; used for the stiffness indicators.
  • standard math The graph contiguity MIP formulation from [42] enforces connectivity.
    Section V.C, used to constrain valid configurations to connected subgraphs.
  • domain assumption Unmodeled aerodynamic forces and moments scale benignly with module count.
    Section II.C.2 claims by dimensional analysis that rotational-velocity-induced effects do not grow with n, but wake interactions are excluded from the dynamics and later measured as a 14 percent hover penalty.
  • domain assumption Distinct lattice points from the restricted connection set place dodecahedra without unintended frame collisions.
    Remark after Fact II.4, stated without proof; needed to allow unconstrained placement of modules at any lattice position in the optimization.
invented entities (1)
  • Dodecacopter module independent evidence
    purpose: Self-contained dodecahedral rotorcraft unit that connects to identical modules to form reconfigurable vehicles.
    The prototype in Section VI was physically built and flown in six configurations, providing direct falsifiable evidence for the module concept.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The Dodecacopter: a Versatile Multirotor System of Dodecahedron-Shaped Modules." pith.science (2026). https://pith.science/paper/WOZBDU7Y

@misc{pith2026250416475,
  author       = {Pith},
  title        = {Pith review of: The Dodecacopter: a Versatile Multirotor System of Dodecahedron-Shaped Modules},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WOZBDU7Y}},
  note         = {Machine review of arXiv:2504.16475}
}
read the original abstract

With the promise of greater safety and adaptability, modular reconfigurable uncrewed air vehicles have been proposed as unique, versatile platforms holding the potential to replace multiple types of monolithic vehicles at once. State-of-the-art rigidly assembled modular vehicles are generally two-dimensional configurations in which the rotors are coplanar and assume the shape of a "flight array". We introduce the Dodecacopter, a new type of modular rotorcraft where all modules take the shape of a regular dodecahedron, allowing the creation of richer sets of configurations beyond flight arrays. In particular, we show how the chosen module design can be used to create three-dimensional and fully actuated configurations. We justify the relevance of these types of configurations in terms of their structural and actuation properties with various performance indicators. Given the broad range of configurations and capabilities that can be achieved with our proposed design, we formulate tractable optimization programs to find optimal configurations given structural and actuation constraints. Finally, a prototype of such a vehicle is presented along with results of performed flights in multiple configurations.

Figures

Figures reproduced from arXiv: 2504.16475 by the authors.

Figure 2
Figure 2. Connection between modules. Connecting faces are [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 4
Figure 4. Examples of conventional configurations recreated with [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figure 5
Figure 5. Tilt angles inducing a positive thrust allowed by the [PITH_FULL_IMAGE:figures/full_fig_p005_5.png] view at source ↗
Figures from the paper (12 more)
Figure 6
Figure 6. Figure 6: Simplified block diagram of a controller for multirotor [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]
Figure 7
Figure 7. Figure 7: 6DOF 12-rotor configuration. Since the power consumption of that module is proportional to |ui | 3 2 , where the absolute value is to ensure the validity of the expression, the average power consumption Pi of module i is the absolute moment of order 3 2 of ui : Pi ∝ E …
Figure 9
Figure 9. Figure 9: Maximum module average power under allocation [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]
Figure 10
Figure 10. Figure 10: Top three rows: Projections of the sets of reachable [PITH_FULL_IMAGE:figures/full_fig_p010_10.png]
Figure 11
Figure 11. Figure 11: Representation of a beam being deformed under the [PITH_FULL_IMAGE:figures/full_fig_p011_11.png]
Figure 12
Figure 12. Figure 12: Deformed quadrotor configuration under external [PITH_FULL_IMAGE:figures/full_fig_p013_12.png]
Figure 13
Figure 13. Figure 13: Deformed configuration under external forces yielding [PITH_FULL_IMAGE:figures/full_fig_p013_13.png]
Figure 14
Figure 14. Figure 14: Module prototype. This problem is a convex quadratic one, whose optimal value is equal to ∥K+c∥2 and reached for v = (K+) 2 c ∥K+c∥2 . Equation (40) is then replaced by Kw = c (43) ∥w∥2 ≤ λ, (44) where w is a newly introduced displacement variable, which, when verifyi…
Figure 15
Figure 15. Figure 15: Three corners belonging to different modules forming [PITH_FULL_IMAGE:figures/full_fig_p016_15.png]
Figure 17
Figure 17. Figure 17: Top view representation of the tetrahedron quadrotor [PITH_FULL_IMAGE:figures/full_fig_p017_17.png]
Figure 18
Figure 18. Figure 18: fig. 18. Vehicles with rotor overlap required a higher average [PITH_FULL_IMAGE:figures/full_fig_p017_18.png]
Figure 18
Figure 18. Figure 18: Control inputs and attitude during hover. [PITH_FULL_IMAGE:figures/full_fig_p018_18.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

57 extracted references · 38 canonical work pages

  1. [1]

    Modular reconfigurable robotics,

    J. Seo, J. Paik, and M. Yim, “Modular reconfigurable robotics,” Annual Review of Control, Robotics, and Autonomous Systems , vol. 2, no. 1, pp. 63–88, 2019. DOI: 10.1146/annurev-control-053018-023834

    work page doi:10.1146/annurev-control-053018-023834 2019

  2. [2]

    The distributed flight array,

    R. Oung and R. D’Andrea, “The distributed flight array,” Mechatronics, vol. 21, no. 6, pp. 908–917, 2011. DOI: 10.1016/j.mechatronics.2010.08.003

    work page doi:10.1016/j.mechatronics.2010.08.003 2011

  3. [3]

    A class of mod- ular aerial robots,

    R. Naldi, F. Forte, and L. Marconi, “A class of mod- ular aerial robots,” in 50th IEEE CDC & ECC , 2011, pp. 3584–3589. DOI: 10.1109/CDC.2011.6161298

    arXiv 2011

  4. [4]

    The LIFT! project – modular, electric vertical lift system with ground power tether,

    M. J. Duffy and A. Samaritano, “The LIFT! project – modular, electric vertical lift system with ground power tether,” in 33rd AIAA AAC, 2015, pp. 3013–3032. DOI: doi/10.2514/6.2015-3013

    work page doi:10.2514/6.2015-3013 2015

  5. [5]

    ModQuad: The flying modular structure that self- assembles in midair,

    D. Saldana, B. Gabrich, G. Li, M. Yim, and V . Kumar, “ModQuad: The flying modular structure that self- assembles in midair,” in IEEE ICRA , 2018, pp. 691–

    work page 2018

  6. [6]

    Universal flying objects: Modular multirotor system for flight of rigid objects,

    B. Mu and P. Chirarattananon, “Universal flying objects: Modular multirotor system for flight of rigid objects,” IEEE Transactions on Robotics, vol. 36, no. 2, pp. 458– 471, 2019. DOI: 10.1109/TRO.2019.2954679

    arXiv 2019

  7. [7]

    Self-reconfiguration in response to faults in modular aerial systems,

    N. Gandhi, D. Saldana, V . Kumar, and L. T. X. Phan, “Self-reconfiguration in response to faults in modular aerial systems,” IEEE Robotics and Automation Letters, vol. 5, no. 2, pp. 2522–2529, 2020. DOI: 10.1109/LRA. 2020.2970685

    arXiv 2020

  8. [8]

    Modeling and experimental validation of a fractal tetrahedron UAS assembly,

    K. Garanger, J. Epps, and E. Feron, “Modeling and experimental validation of a fractal tetrahedron UAS assembly,” in IEEE AeroConf, Big Sky, MT, USA, 2020, pp. 1–11. DOI: 10.1109/AERO47225.2020.9172614

    arXiv 2020

Show all 57 references
  1. [9]

    The reconfig- urable aerial robotic chain: Shape and motion planning,

    M. Kulkarni, H. Nguyen, and K. Alexis, “The reconfig- urable aerial robotic chain: Shape and motion planning,” IFAC-PapersOnLine, vol. 53, no. 2, pp. 9295–9302,

  2. [10]

    Reconfigurable drone system for transportation of parcels with variable mass and size,

    F. Schiano, P. M. Kornatowski, L. Cencetti, and D. Flo- reano, “Reconfigurable drone system for transportation of parcels with variable mass and size,” IEEE Robotics and Automation Letters , vol. 7, no. 4, pp. 12 150– 12 157, 2022. DOI: 10.1109/LRA.2022.3208716

    2022

  3. [11]

    Mod- ular reconfigurable rotorcraft for autonomous cargo delivery,

    J. Davis, M. Park, I. Shapiro, and M. Costello, “Mod- ular reconfigurable rotorcraft for autonomous cargo delivery,” Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering , vol. 238, no. 14, pp. 1419–1439, 2024. DOI: 10.1177/ 09544100...

    2024

  4. [12]

    Modular, teth- ered, reconfigurable hybrid robotic vehicles for aerial grasping and package delivery,

    M. Kobayashi, G. Huang, J. Buzzatto, A. Lynch, K. Kyriakopoulos, and M. Liarokapis, “Modular, teth- ered, reconfigurable hybrid robotic vehicles for aerial grasping and package delivery,” in IEEE SSRR , 2024, pp. 131–135. DOI: 10 . 1109 / SSRR62954 . 2024 . 10770002

    2024

  5. [13]

    Modeling and switched control of modular reconfigurable flight array for faulty redundancy,

    B. Ren, C. Yang, X. Zhang, and W. Mao, “Modeling and switched control of modular reconfigurable flight array for faulty redundancy,” Processes, vol. 12, no. 4,

  6. [14]

    Fundamental actuation properties of multirotors: Force–moment de- coupling and fail–safe robustness,

    G. Michieletto, M. Ryll, and A. Franchi, “Fundamental actuation properties of multirotors: Force–moment de- coupling and fail–safe robustness,” IEEE Transactions on Robotics , vol. 34, no. 3, pp. 702–715, 2018. DOI: 10.1109/TRO.2018.2821155

    2018

  7. [15]

    Modeling, control and design optimization for a fully- actuated hexarotor aerial vehicle with tilted propellers,

    S. Rajappa, M. Ryll, H. H. B ¨ulthoff, and A. Franchi, “Modeling, control and design optimization for a fully- actuated hexarotor aerial vehicle with tilted propellers,” in IEEE ICRA , 2015, pp. 4006–4013. DOI: 10.1109/ ICRA.2015.7139759

    2015

  8. [16]

    Design, modeling and control of an omni-directional aerial vehicle,

    D. Brescianini and R. D’Andrea, “Design, modeling and control of an omni-directional aerial vehicle,” in IEEE ICRA, 2016, pp. 3261–3266

    2016

  9. [17]

    A maneuverability analysis of a novel hexarotor UA V concept,

    H. Mehmood, T. Nakamura, and E. N. Johnson, “A maneuverability analysis of a novel hexarotor UA V concept,” in IEEE ICUAS , 2016, pp. 437–446. DOI: 10.1109/ICUAS48674.2020.9214004

    2016

  10. [18]

    Maneuver- ability analysis of a fully-actuated hexrotor UA V con- sidering tilt angles and arrangement of rotors,

    Y . Tadokoro, T. Ibuki, and M. Sampei, “Maneuver- ability analysis of a fully-actuated hexrotor UA V con- sidering tilt angles and arrangement of rotors,” IFAC- PapersOnLine, vol. 50, no. 1, pp. 8981–8986, 2017. DOI: 10.1016/j.ifacol.2017.08.1325

    2017 doi

  11. [19]

    Past, present, and future of aerial robotic manipulators,

    A. Ollero, M. Tognon, A. Suarez, D. Lee, and A. Franchi, “Past, present, and future of aerial robotic manipulators,” IEEE Transactions on Robotics, vol. 38, no. 1, pp. 626–645, 2021. DOI: 10.1109/TRO.2021. 3084395

    2021 doi

  12. [20]

    Design, modeling and control of omni-directional aerial robot,

    S. Park, J. Her, J. Kim, and D. Lee, “Design, modeling and control of omni-directional aerial robot,” in IEEE IROS, 2016, pp. 1570–1575. DOI: 10.1109/IROS.2016. 7759254

    2016 doi

  13. [21]

    The V oliro omniorientational hexacopter: An agile and maneuver- able tiltable-rotor aerial vehicle,

    M. Kamel, S. Verling, O. Elkhatib, et al. , “The V oliro omniorientational hexacopter: An agile and maneuver- able tiltable-rotor aerial vehicle,” IEEE Robotics & Automation Magazine, vol. 25, no. 4, pp. 34–44, 2018. DOI: 10.1109/MRA.2018.2866758. 20

    2018

  14. [22]

    A novel robotic platform for aerial manipulation using quadro- tors as rotating thrust generators,

    H.-N. Nguyen, S. Park, J. Park, and D. Lee, “A novel robotic platform for aerial manipulation using quadro- tors as rotating thrust generators,” IEEE Transactions on Robotics , vol. 34, no. 2, pp. 353–369, 2018. DOI: 10.1109/TRO.2018.2791604

    2018

  15. [23]

    Design and optimal control of a tiltrotor micro-aerial vehicle for efficient omnidirectional flight,

    M. Allenspach, K. Bodie, M. Brunner, et al. , “Design and optimal control of a tiltrotor micro-aerial vehicle for efficient omnidirectional flight,” The International Jour- nal of Robotics Research, vol. 39, no. 10-11, pp. 1305– 1325, 2020. DOI: 10.1177/0278364920943654

    2020 doi

  16. [24]

    ModQuad-DoF: A novel yaw actuation for modular quadrotors,

    B. Gabrich, G. Li, and M. Yim, “ModQuad-DoF: A novel yaw actuation for modular quadrotors,” in IEEE ICRA, 2020, pp. 8267–8273. DOI: 10.1109/ICRA40945. 2020.9196735

    2020

  17. [25]

    Wake interactions of a tetrahedron quadcopter,

    J. Epps, K. Garanger, and E. Feron, “Wake interactions of a tetrahedron quadcopter,” in IEEE ICUAS , 2020, pp. 1852–1859

    2020

  18. [26]

    Classifying the finite subgroups of so (3)

    H. T. A. Bui. “Classifying the finite subgroups of so (3).” (2020), [Online]. Available: https://math.uchicago. edu/%7Emay/REU2020/REUPapers/Bui,An.pdf

    2020

  19. [27]

    H. S. M. Coxeter, Regular Polytopes, 3rd. Mineola, NY , USA: Dover Publications, 1973

    1973

  20. [28]

    Modelling and control of quadcopter,

    T. Luukkonen, “Modelling and control of quadcopter,” School of Science, Aalto University, Tech. Rep., 2011. [Online]. Available: https://sal.aalto.fi/publications/pdf- files/eluu11 public.pdf

    2011

  21. [29]

    Accurate tracking of aggres- sive quadrotor trajectories using incremental nonlin- ear dynamic inversion and differential flatness,

    E. Tal and S. Karaman, “Accurate tracking of aggres- sive quadrotor trajectories using incremental nonlin- ear dynamic inversion and differential flatness,” IEEE Transactions on Control Systems Technology , vol. 29, no. 3, pp. 1203–1218, 2020. DOI: 10.1109/TCST.2020. 3001117

    2020 doi

  22. [30]

    Differ- ential flatness of quadrotor dynamics subject to rotor drag for accurate tracking of high-speed trajectories,

    M. Faessler, A. Franchi, and D. Scaramuzza, “Differ- ential flatness of quadrotor dynamics subject to rotor drag for accurate tracking of high-speed trajectories,” IEEE Robotics and Automation Letters , vol. 3, no. 2, pp. 620–626, 2017. DOI: 10.1109/LRA.2017.2776353

    2017

  23. [31]

    Boyd and L

    S. Boyd and L. Vandenberghe, Convex Optimization . Cambridge, United Kingdom: Cambridge University Press, 2004. [Online]. Available: https://web.stanford. edu/∼boyd/cvxbook

    2004

  24. [32]

    Computing the maximum volume inscribed ellipsoid of a polytopic projection,

    J. Zhen and D. Den Hertog, “Computing the maximum volume inscribed ellipsoid of a polytopic projection,” INFORMS Journal on Computing , vol. 30, no. 1, pp. 31–42, 2018. DOI: 10.5555/3215359.3215362

    2018

  25. [33]

    Solving control allocation problems using semidefinite programming,

    J.-H. Oh, M. B. Jamoom, M. W. McConley, and E. Feron, “Solving control allocation problems using semidefinite programming,” Journal of Guidance, Con- trol, and Dynamics , vol. 22, no. 3, pp. 494–497, 1999. DOI: 10.2514/2.7633

    1999 doi

  26. [34]

    G. J. Leishman, Principles of Helicopter Aerodynamics, 2nd. Cambridge University Press, 2016

    2016

  27. [35]

    Blade element mo- mentum modeling of low-Re small UAS electric propul- sion systems,

    M. McCrink and J. W. Gregory, “Blade element mo- mentum modeling of low-Re small UAS electric propul- sion systems,” in 33rd AIAA AAC, 2015, pp. 3296–3318. DOI: 10.2514/6.2015-3296

    2015 doi

  28. [36]

    Moments and absolute moments of the normal distribution

    A. Winkelbauer. “Moments and absolute moments of the normal distribution.” version 2. (2012), [Online]. Available: https://arxiv.org/abs/1209.4340

    2012 arXiv

  29. [37]

    Discussion and practical as- pects on control allocation for a multi-rotor helicopter,

    G. Ducard and M.-D. Hua, “Discussion and practical as- pects on control allocation for a multi-rotor helicopter,” in ISPRS UAV-g, vol. 38, 2011, pp. 95–100. DOI: 10. 5194/isprsarchives-XXXVIII-1-C22-95-2011

    2011

  30. [38]

    Dynamic control allocation using con- strained quadratic programming,

    O. H ¨arkeg˚ard, “Dynamic control allocation using con- strained quadratic programming,” Journal of Guidance, Control, and Dynamics , vol. 27, no. 6, pp. 1028–1034,

  31. [39]

    Energy optimal control allocation in a redundantly actuated omnidirectional UA V,

    E. Dyer, S. Sirouspour, and M. Jafarinasab, “Energy optimal control allocation in a redundantly actuated omnidirectional UA V,” inIEEE ICRA, 2019, pp. 5316–

    2019

  32. [40]

    Kassimali, Matrix Analysis of Structures , 2nd

    A. Kassimali, Matrix Analysis of Structures , 2nd. Boston, MA, USA: Cengage Learning, 2012

    2012

  33. [41]

    Weaver and J

    W. Weaver and J. M. Gere, Matrix Analysis of Framed Structures, 3rd. Berlin, Germany: Springer, 1990, ISBN : 978-1-4684-7487-9

    1990

  34. [42]

    A model of contiguity for spatial unit allo- cation,

    T. Shirabe, “A model of contiguity for spatial unit allo- cation,” Geographical Analysis, vol. 37, no. 1, pp. 2–16,

  35. [43]

    F. S. Hillier and G. J. Lieberman, Introduction to Operations Research, 7th ed. McGraw-Hill, 2022

    2022

  36. [44]

    Diestel, Graph Theory , 6th ed

    R. Diestel, Graph Theory , 6th ed. Springer Berlin, Heidelberg, 2025

    2025

  37. [45]

    Y ALMIP : A Toolbox for Modeling and Optimization in MATLAB,

    J. L ¨ofberg, “Y ALMIP : A Toolbox for Modeling and Optimization in MATLAB,” in CACSD, 2004, pp. 284–

    2004

  38. [46]

    [Online]

    Gurobi Optimization, LLC, Gurobi optimizer reference manual, 2022. [Online]. Available: https://www.gurobi. com

    2022

  39. [47]

    [Online]

    ArduPilot Dev Team, ArduPilot documentation , 2022. [Online]. Available: https://ardupilot.org/ardupilot

    2022

  40. [48]

    Controlling an assembly of aerial vehicles under uncertainty using an adaptive switching approach,

    M. T. Shahab, K. Garanger, and E. Feron, “Controlling an assembly of aerial vehicles under uncertainty using an adaptive switching approach,” in AACC ACC, 2022, pp. 514–519. DOI: 10.23919/ACC53348.2022.9867535

    2022

  41. [49]

    An empirical analysis of the wake interactions of a novel multi-agent UAS,

    J. T. Epps, K. Garanger, T. Khamvilai, and E. Feron, “An empirical analysis of the wake interactions of a novel multi-agent UAS,” in AIAA SCITECH, 2022. DOI: 10.2514/6.2022-0280

    2022 doi

  42. [50]

    [Online]

    NIST Digital Library of Mathematical Functions, Olver, F W J and Olde Daalhuis, A B and Lozier D W and Schneider, B I and Boisvert, R F and Clark, C W and Miller, B R and Saunders, B V and Cohl, H S and McClain, M A. [Online]. Available: https://dlmf.nist. gov/

  43. [289]

    DOI: 10.1109/CACSD.2004.1393890

    2004 arXiv

  44. [698]

    DOI: 10.1109/ICRA.2018.8461014

    2018

  45. [2004]

    DOI: 10.2514/1.11607

    doi

  46. [2005]

    DOI: 10.1111/j.1538-4632.2005.00605.x

    2005

  47. [2020]

    DOI: doi.org/10.1016/j.ifacol.2020.12.2383

    2020 doi

  48. [2024]

    DOI: 10.3390/pr12040646

    doi

  49. [5322]

    DOI: 10.1109/ICRA.2019.8793549

    2019

Pith tools

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