Pith. sign in

REVIEW 4 major objections 4 minor 2 cited by

By planning in a flat-output space where the dynamics are linear, this paper obtains closed-form polynomial local paths that convert back to dynamically feasible robot trajectories, making sampling-based kinodynamic planning fast enough for

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-02 18:03 UTC pith:LEL7FYL5

load-bearing objection Useful integration for r=2 systems; quadrotor claims overreach and contradict the paper's own flatness orders. the 4 major comments →

arxiv 2603.16059 v3 pith:LEL7FYL5 submitted 2026-03-17 cs.RO

Ultrafast Sampling-based Kinodynamic Planning via Differential Flatness

Thai Duong , Clayton W. Ramsey , Zachary Kingston , Wil Thomason , Lydia E. Kavraki This is my paper
classification cs.RO
keywords kinodynamic planningdifferential flatnesssampling-based motion planningclosed-form boundary value problemSIMD parallel collision checkingprobabilistic exhaustivityasymptotic optimality
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The paper tries to show that the hardest part of kinodynamic planning—solving two-point boundary value problems under nonlinear dynamics—can be sidestepped for differentially flat robots. Because the flat output and its derivatives determine the full state and control, planning can be moved to a linear flat state where the boundary value problem has a closed-form polynomial solution. Those polynomials can be converted back into dynamically feasible trajectories and checked for collisions simultaneously with SIMD CPU instructions. If this holds, sampling-based planners that normally plan geometric paths can become kinodynamic planners with microsecond-to-millisecond runtimes, without giving up asymptotic optimality.

Core claim

The central claim is that differential flatness converts the kinodynamic planning problem into planning in a flat state space with linear dynamics, where the linear-quadratic minimum-time boundary value problem has a closed-form solution: the optimal pseudo-control is a polynomial and the flat trajectory is a piecewise polynomial. Mapping these polynomials back through the flatness equations yields dynamically feasible state trajectories by construction, so collision checking can be done in parallel on sampled states. The paper further claims probabilistic exhaustivity and asymptotic optimality for the resulting planners, with convergence bounds decaying as O((log N / N)^(1/D)).

What carries the argument

The enabling object is the flat-state representation z = (y, y_dot, ..., y^(r-1)) with pseudo-control w = y^(r), which turns the nonlinear dynamics into a linear chain-of-integrators system with a nilpotent matrix A. The linear-quadratic minimum-time BVP solution uses the Gramian G_T to give closed-form polynomial local paths, and the nilpotence of A makes e^{At} and G_T polynomials in time. These polynomials are what allow dense time-discretization and SIMD-parallel collision checking.

Load-bearing premise

The load-bearing premise is that planning in the flat space with the chosen number of derivatives is exactly equivalent to planning in the original state space; for quadrotors, the paper uses fewer derivatives than a direct reconstruction of the full state would need, so that equivalence is the first assumption to verify.

What would settle it

Take a quadrotor local path generated by the r=2 formula, reconstruct the rotation matrix and angular velocity from the flat-output trajectory at several sample times, and verify the full dynamics against the returned thrust and torque; a violation, or a planned path through a narrow gap whose reconstructed attitude exceeds the vehicle's physical limits, would show that the returned trajectory is not a complete dynamically feasible quadrotor state trajectory.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Any sampling-based planner that uses a CONNECT/EXTEND loop (e.g., RRT-Connect or SST*) can be turned into a kinodynamic planner while keeping trajectories dynamically feasible by construction.
  • Planning times drop to the microsecond-to-millisecond range on commodity CPUs, making online reactive replanning in dynamic environments practical; the paper demonstrates about 90 microseconds of planning in a real pick-and-place loop.
  • The closed-form BVP solutions allow shortcut-based trajectory simplification, reducing the wandering behavior common in propagation-based kinodynamic planners.
  • If the optimality analysis is correct, existing asymptotic optimality guarantees of the host sampling planner are preserved, not broken, by the flat-space reformulation.
  • Trajectory postprocessing with BVP shortcuts shortens trajectories without recomputing dynamics numerically.

Where Pith is reading between the lines

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

  • The speed argument depends on polynomial local paths staying polynomial under the recovery maps alpha and beta; for systems where those maps involve transcendental functions such as arctan2, the closed-form polynomial property may only hold for a restricted set of outputs, which would limit the 'broad class' claim.
  • A concrete test of the quadrotor claim: generate a local path with r=2 and reconstruct R and omega from the flat state; if the reconstructed attitude and angular velocity do not satisfy the original dynamics with the returned thrust and torque, then dynamic feasibility for quadrotors is only established for the positional substate, not the full state.
  • The probabilistic exhaustivity proof assumes alpha is Lipschitz-continuous, but for rotational states on SO(3) Lipschitz continuity is only local; a global statement would require charts or a metric on SO(3), which the paper leaves implicit.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

4 major / 4 minor

Summary. The paper proposes FLASK, a sampling-based kinodynamic planning framework for differentially flat robot systems. The key idea is to map the planning problem to the flat output space, where the dynamics become a linear nilpotent system. The authors derive a closed-form solution to the linear-quadratic minimum-time boundary value problem (Eqs. 23-25), which yields time-parameterized polynomial local paths. These paths are amenable to SIMD-parallelized collision checking, and the resulting trajectory is converted back to the original state space via the flatness maps α and β. The paper claims this framework is exact, applies to manipulators, ground vehicles, aerial vehicles, and more, and provides probabilistic exhaustivity and asymptotic optimality guarantees. Experiments on unicycles, 2D/3D quadrotors, a 7-DOF Franka manipulator, and a real UR5 platform show planning times in microseconds to milliseconds.

Significance. If the claims are correct, this would be a meaningful advance: it would show that SIMD-accelerated geometric planning techniques can be extended to kinodynamic planning for a broad class of nonlinear systems, with closed-form BVP solutions and theoretical guarantees. The manipulator and unicycle results, and the integration with VAMP/CAPT-style SIMD collision checking, are promising and clearly presented. The closed-form derivation for r=2 systems is clean and follows known LQMT results. However, the aerial-vehicle portion of the paper has a correctness gap that undermines the broad-class claim, and the theoretical analysis is a sketch rather than a complete proof. The contribution is significant but requires substantial revision before it can be accepted.

major comments (4)
  1. [Sec. V-C1b and Example 3 / Eq. (14)] The statement that quadrotors 'can reduce the pseudo-control order to r=2' directly contradicts Eq. (14), which requires r ≥ max(l,m). Example 3 establishes l=3 (state recovery needs p^(3)) and m=4 (control needs p^(4)). With r=2, the flat state z=(p,ψ,ṗ,ψ̇) and pseudo-control w=(p̈,ψ̈) do not contain the derivatives needed by α and β in Eqs. (11)-(13) to recover R, ω, f, and τ. Consequently the trajectories generated for 2D and 3D quadrotors in Sec. VII-A are not verified dynamically feasible in the original state space, and the stated limits on ω and τ cannot be checked. This also violates Assumption 1 (r>l) in Theorem 1, so the theoretical guarantees do not cover the quadrotor experiments. The broad-class claim including aerial vehicles is therefore unsupported as stated.
  2. [Sec. VI-A, Theorem 1 proof] The proof of probabilistic exhaustivity invokes Theorem IV.6 of [103] but does not verify that the graph produced by the actual FLASK algorithm (Alg. 1 + Alg. 2) meets that theorem's hypotheses. Specifically, it must be shown that (i) the optimal-time local path in Eq. (25) has the required cost/ellipsoid scaling and Lipschitz properties, (ii) the node-selection rule in Alg. 2 (cost threshold ζ) accepts the needed connections with high probability, and (iii) the collision-checking step does not reject the connecting paths with probability that destroys the guarantee. The proof only establishes bounds on ellipsoid volumes and then jumps to the external theorem; it does not connect the random geometric graph to the algorithm's sampling and extension procedure. As written, Eqs. (33)-(34) are not a proof for FLASK as specified.
  3. [Sec. VI-B, Optimality analysis] The asymptotic optimality argument assumes 'there exists a sequence of δ_N-clearance trajectories σ̂_z^(N)(t) with cost converging to C*'. This is an additional regularity assumption on the optimal trajectory that is not derived or discussed. More importantly, the text says the bound ζ_N 'can be used as thresholds for expanding and rewiring the graph' but does not demonstrate that the actual planners used (FLASK-RRTConnect, FLASK-SST*) perform the rewiring or dynamic-programming steps necessary for the standard optimality proof. The claim that the framework 'preserves asymptotic-optimality guarantees' of any sampling-based planner is therefore not substantiated; at most it is an outline of how a proof could proceed.
  4. [Sec. V-B, Problem 2 equivalence] The equivalence between Problem 1 and Problem 2 hinges on the maps α and β being well-defined on the flat state. For the systems where r≥max(l,m) in Eq. (14), this is mathematically sound. However, for the quadrotor case actually used in the experiments (r=2), the equivalence fails: Problem 2 requires u(t)=β(σ_z(t),w(t))∈U, but β is not defined when the required derivatives are absent. The paper should either restrict the framework to systems with r≥max(l,m), or re-derive the quadrotor experiments using r=4 (or a trajectory representation that retains p^(3), p^(4)).
minor comments (4)
  1. [General / heading] Typo in Sec. VI heading: 'Probabilistic Exhausitivity' should be 'Exhaustivity'. Also in the introduction, 'high-DOF' is written as 'high-D OF'.
  2. [Sec. V-C1b] The citation to [21] for the claim that quadrotors can use r=2 should be rechecked; [21] plans with acceleration-level controls for a simplified model and does not provide a full flatness-based state recovery with r=2. This citation does not resolve the contradiction.
  3. [Table I] In the 3D quadrotor rows, the environment labeled 'Block' appears to correspond to the 'Obstacle' environment in Fig. 5(c). Please make the naming consistent.
  4. [Alg. 3 / notation] In Alg. 3, the loop bound 'N/K' should likely be '⌈N/K⌉-1' and the batch definition uses ⌈N/K⌉; please clarify the indexing so the total number of samples is exactly N.

Circularity Check

0 steps flagged

No significant circularity; the quadrotor r=2 reduction is a correctness gap, not circular derivation.

full rationale

The paper's central derivation is self-contained and not circular. The transformation from Problem 1 to the flat-state linear system (Eqs. 14-17) is a coordinate change justified by the standard definition of differential flatness (Def. 1), not a fit that is later renamed as a prediction. The closed-form BVP solution (Eqs. 23-25) follows from the known linear-quadratic minimum-time solution [94], and the conversion back to the original state space (Eqs. 19, 30) is exactly the flatness mapping. The probabilistic exhaustivity proof (Theorem 1) is an adaptation of external random geometric graph theory [103, 104] to the nilpotent linear structure of Eq. (16); it does not assume the conclusion. Self-citations such as VAMP [7], CAPT [8], and MotionBenchMaker [107] are engineering and benchmarking dependencies, not load-bearing assumptions that reduce the proof to the paper's own claims. The most serious issue in the manuscript -- the statement that quadrotors can reduce the pseudo-control order to r=2 (Sec. V-C1b) despite Example 3 requiring l=3, m=4 and Eq. (14) requiring r >= max(l,m)=4 -- is an internal consistency and correctness problem for the aerial-vehicle experiments, but it is not a circular derivation: no parameter is fitted and then 'predicted,' and no load-bearing conclusion is justified solely by an author self-citation. Therefore, per the hard rules, the circularity score is 0.

Axiom & Free-Parameter Ledger

6 free parameters · 6 axioms · 0 invented entities

The method introduces no new physical entity. It relies on standard differential-flatness conversions, a linear flat-state reformulation, a closed-form LQMT BVP solution, and geometric random-graph arguments. The main uncharged premises are the derivative-order condition r>l, Lipschitzness of alpha, existence of clearance-approximating optimal trajectories, and finite-sample constraint validation. The quadrotor r=2 reduction conflicts with the paper's own derivative-order bookkeeping.

free parameters (6)
  • rho (time/control cost weight) = rho=1 (experiments)
    Trade-off parameter in cost (20) balancing control effort against duration; affects minimum-time T* via (26), set by hand.
  • R weight matrix = R=I_n (experiments)
    Positive definite weight in LQMT cost (20); user-specified.
  • Cost threshold zeta for choosing existing node = not specified
    FLASKEXTEND uses Cloc < zeta to pick z0; theorem only links zeta to N asymptotically, algorithm gives no concrete schedule.
  • Number of collision-checking samples N in Algorithm 3 = not specified
    Collision/constraint validation is done on discrete samples; N controls the sampling-resolution guarantee and is not stated.
  • Clearance sequence delta_N = delta_N=O(eta_N) (suggested)
    Optimality outline requires delta_N-clearance trajectories approximating the optimum; choice is asymptotic only, not constructive.
  • kappa in Theorem 1 = kappa>=0, unspecified
    User-defined parameter in probability bound (32)-(33).
axioms (6)
  • domain assumption The robot system is differentially flat with known alpha, beta (Def. 1).
    The entire method is restricted to systems with valid flat outputs; examples rely on standard catalogs [34,35].
  • ad hoc to paper r >= max(l,m) in Eq. (14), and Assumption 1 requires r>l so x=alpha(z).
    Needed for Problem 1/2 equivalence and Theorem 1(ii). The paper's own quadrotor discussion using r=2 contradicts this for Example 3, where l=3.
  • domain assumption The conversion alpha is Lipschitz-continuous on Z.
    Theorem 1(ii) transfers exhaustivity bounds to the original state space; not proven for manipulator/unicycle/quadrotor examples.
  • ad hoc to paper There exists a sequence of delta_N-clearance trajectories sigma_hat_z^(N) with cost -> C* (Sec. VI-B).
    Unproven regularity/topological assumption; used to turn local exhaustivity into asymptotic optimality.
  • standard math External results: LQMT closed-form solution [94], controllability Gramian scaling Lemma III.4 and Theorem IV.6 from [103].
    BVP solution and probabilistic exhaustivity coverage argument are imported from prior work; not re-derived.
  • domain assumption State/control limits can be checked by finite sampling of converted (x_loc,u_loc) or bounded in flat space.
    Section V-D; exactness of constraint satisfaction depends on sample density or conservative bounds, which are not fully specified.

pith-pipeline@v1.3.0-alltime-deepseek · 34104 in / 14662 out tokens · 137266 ms · 2026-08-02T18:03:50.396412+00:00 · methodology

0 comments
read the original abstract

Motion planning under dynamics constraints, i.e, kinodynamic planning, enables safe robot operation by generating dynamically feasible trajectories that the robot can accurately track. For high-DOF robots such as manipulators, sampling-based motion planners are commonly used, especially for complex tasks in cluttered environments. However, enforcing constraints on robot dynamics in such planners requires solving either challenging two-point boundary value problems (BVPs) or propagating robot dynamics, both of which cause computational bottlenecks that drastically increase planning times. Meanwhile, recent efforts have shown that sampling-based motion planners can generate plans in microseconds using parallelization, but are limited to geometric paths. This paper develops FLASK, a fast parallelized sampling-based kinodynamic motion planning framework for a broad class of differentially flat robot systems, including manipulators, ground and aerial vehicles, and more. Differential flatness allows us to transform the motion planning problem from the original state space to a flat output space, where an analytical time-parameterized solution of the BVP problem can be obtained. A trajectory in the flat output space is then converted back to a closed-form dynamically feasible trajectory in the original state space, enabling fast validation via ``single instruction, multiple data" parallelism. Our framework is fast, exact, and compatible with any sampling-based motion planner, while offering theoretical guarantees on probabilistic exhaustibility and asymptotic optimality based on the closed-form BVP solutions. We extensively verify the effectiveness of our approach in both simulated benchmarks and real experiments with cluttered and dynamic environments, requiring mere microseconds to milliseconds of planning time.

Figures

Figures reproduced from arXiv: 2603.16059 by Clayton W. Ramsey, Lydia E. Kavraki, Thai Duong, Wil Thomason, Zachary Kingston.

Figure 1
Figure 1. Figure 1: Motion planning for a “pick and place” task in a cluttered environment: a dynamically feasible trajectory (a) generated from our FLASK framework can be accurately tracked by a UR5 robot. Meanwhile, tracking a geometric path (b) leads to collisions (shown in red) that topple the nearby boxes. Multiple intermediate states are overlaid to illustrate the robot’s motion. Our planning framework is real-time and … view at source ↗
Figure 2
Figure 2. Figure 2: Configuration samples a, b, c and d, discretized from a linear path (a), as in VAMP [7], and from our closed-form time-parameterized motions (b), can be efficiently checked for collision using SIMD parallelism. V. TECHNICAL APPROACH In this section, we present our kinodynamic planning frame￾work, FLASK, by showing that the flat output evolves as a linear system (Sec. V-A), and hence, allows us to convert P… view at source ↗
Figure 3
Figure 3. Figure 3: Formulating the kinodynamic planning problem in the flat state space [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Our theoretical analysis: (a) Probabilistic exhaustivity: as [PITH_FULL_IMAGE:figures/full_fig_p011_4.png] view at source ↗
Figure 6
Figure 6. Figure 6: Visualization of trajectories generated by our [PITH_FULL_IMAGE:figures/full_fig_p014_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Our trajectory is smoother and dynamically feasible, leading to better tracking performance. For example, the third joint angle [PITH_FULL_IMAGE:figures/full_fig_p016_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: The “pick and place” task with UR5 robot in a cluttered environment with narrow passages: (a) our kinodynamic planner successfully finishes the task [PITH_FULL_IMAGE:figures/full_fig_p016_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Reactive planning with moving obstacles: our UR5 robot successfully [PITH_FULL_IMAGE:figures/full_fig_p017_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Planning and simplification time in our “p [PITH_FULL_IMAGE:figures/full_fig_p017_10.png] view at source ↗

Review history (2 revisions) →

discussion (0)

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

Forward citations

Cited by 2 Pith papers

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

  1. Compliant Sphere Lattice Contact: Distributed Contact Modeling for Sphere-Based Robot Representations

    cs.RO 2026-07 conditional novelty 6.0

    CSLC models sphere-based robot surfaces as a spring lattice, producing distributed contact patches and restoring torque that point contact lacks.

  2. Conformal Constraint Tightening for Chance-Constrained Motion Planning with Unknown Dynamics

    cs.RO 2026-07 reject novelty 6.0

    A conformal-prediction bound on nominal-to-true trajectory deviation is used to tighten motion-planning constraints, yielding a probabilistic task-completion guarantee under model mismatch.

Reference graph

Works this paper leans on

108 extracted references · cited by 2 Pith papers

  1. [1]

    A Re- view of Motion Planning for Highway Autonomous Driving

    L. Claussmann, M. Revilloud, D. Gruyer, and S. Glaser. “A Re- view of Motion Planning for Highway Autonomous Driving”. In: IEEE Transactions on Intelligent Transportation Systems21.5 (2020), pp. 1826–1848

    2020

  2. [2]

    Trajectory Planning for Quadrotor Swarms

    W. H¨onig, J. A. Preiss, T. K. S. Kumar, G. S. Sukhatme, and N. Ayanian. “Trajectory Planning for Quadrotor Swarms”. In:IEEE Transactions on Robotics34.4 (2018), pp. 856–869

    2018

  3. [3]

    Swarm of micro flying robots in the wild

    X. Zhou, X. Wen, Z. Wang, Y . Gao, H. Li, Q. Wang, T. Yang, H. Lu, Y . Cao, C. Xu, et al. “Swarm of micro flying robots in the wild”. In: Science Robotics7.66 (2022)

    2022

  4. [4]

    Lessons from the amazon picking challenge: Four aspects of building robotic systems

    C. Eppner, S. H¨ofer, R. Jonschkowski, R. Mart´ın-Mart´ın, A. Sieverling, V . Wall, and O. Brock. “Lessons from the amazon picking challenge: Four aspects of building robotic systems.” In:Robotics: Science and Systems. 2016

    2016

  5. [5]

    Healthcare robotics

    L. D. Riek. “Healthcare robotics”. In:Communications of the ACM 60.11 (2017), pp. 68–78

    2017

  6. [6]

    FEAST: A Flexible Mealtime-Assistance System Towards In-the-Wild Personalization

    R. K. Jenamani, T. Silver, B. Dodson, S. Tong, A. Song, Y . Yang, Z. Liu, B. Howe, A. Whitneck, and T. Bhattacharjee. “FEAST: A Flexible Mealtime-Assistance System Towards In-the-Wild Personalization”. In:Robotics: Science and Systems. Los Angeles, CA, USA, 2025

    2025

  7. [7]

    Motions in Microsec- onds via Vectorized Sampling-Based Planning

    W. Thomason, Z. Kingston, and L. E. Kavraki. “Motions in Microsec- onds via Vectorized Sampling-Based Planning”. In:IEEE Interna- tional Conference on Robotics and Automation. 2024, pp. 8749–8756

    2024

  8. [8]

    Collision-Affording Point Trees: SIMD-Amenable Nearest Neighbors for Fast Collision Checking

    C. W. Ramsey, Z. Kingston, W. Thomason, and L. E. Kavraki. “Collision-Affording Point Trees: SIMD-Amenable Nearest Neighbors for Fast Collision Checking”. In:Robotics: Science and Systems. 2024

    2024

  9. [9]

    Randomized Kinodynamic Planning

    S. M. LaValle and J. J. K. Jr. “Randomized Kinodynamic Planning”. In: The International Journal of Robotics Research20.5 (2001), pp. 378– 400

    2001

  10. [10]

    Randomized Kinody- namic Motion Planning with Moving Obstacles

    D. Hsu, R. Kindel, J.-C. Latombe, and S. Rock. “Randomized Kinody- namic Motion Planning with Moving Obstacles”. In:The International Journal of Robotics Research21.3 (2002), pp. 233–255

    2002

  11. [11]

    S. M. LaValle.Planning algorithms. Cambridge University Press, 2006

    2006

  12. [12]

    Generation of collision-free trajectories for a quadrocopter fleet: A sequential convex programming approach

    F. Augugliaro, A. P. Schoellig, and R. D’Andrea. “Generation of collision-free trajectories for a quadrocopter fleet: A sequential convex programming approach”. In:IEEE/RSJ International Conference on Intelligent Robots and Systems. 2012, pp. 1917–1922

    2012

  13. [13]

    Motion planning with sequential convex optimization and convex collision checking

    J. Schulman, Y . Duan, J. Ho, A. Lee, I. Awwal, H. Bradlow, J. Pan, S. Patil, K. Goldberg, and P. Abbeel. “Motion planning with sequential convex optimization and convex collision checking”. In:The Interna- tional Journal of Robotics Research33.9 (2014), pp. 1251–1270

    2014

  14. [14]

    GuSTO: Guaran- teed Sequential Trajectory optimization via Sequential Convex Pro- gramming

    R. Bonalli, A. Cauligi, A. Bylard, and M. Pavone. “GuSTO: Guaran- teed Sequential Trajectory optimization via Sequential Convex Pro- gramming”. In:International Conference on Robotics and Automation. 2019, pp. 6741–6747

    2019

  15. [15]

    Synthesis and stabilization of complex behaviors through online trajectory optimization

    Y . Tassa, T. Erez, and E. Todorov. “Synthesis and stabilization of complex behaviors through online trajectory optimization”. In: IEEE/RSJ International Conference on Intelligent Robots and Systems. 2012, pp. 4906–4913

    2012

  16. [16]

    Crocoddyl: An Efficient and Versatile Framework for Multi-Contact Optimal Control

    C. Mastalli, R. Budhiraja, W. Merkt, G. Saurel, B. Hammoud, M. Naveau, J. Carpentier, L. Righetti, S. Vijayakumar, and N. Mansard. “Crocoddyl: An Efficient and Versatile Framework for Multi-Contact Optimal Control”. In:IEEE International Conference on Robotics and Automation. 2020, pp. 2536–2542

    2020

  17. [17]

    iDb-A*: Iterative Search and Optimization for Optimal Kinodynamic Motion Planning

    J. Ortiz-Haro, W. H ¨onig, V . N. Hartmann, and M. Toussaint. “iDb-A*: Iterative Search and Optimization for Optimal Kinodynamic Motion Planning”. In:IEEE Transactions on Robotics41 (2025)

    2025

  18. [18]

    Efficient constrained path planning via search in state lattices

    M. Pivtoraiko and A. Kelly. “Efficient constrained path planning via search in state lattices”. In:International Symposium on Artificial Intelligence, Robotics, and Automation in Space. 2005, pp. 1–7

    2005

  19. [19]

    Kinodynamic motion planning with state lattice motion primitives

    M. Pivtoraiko and A. Kelly. “Kinodynamic motion planning with state lattice motion primitives”. In:IEEE/RSJ International Conference on Intelligent Robots and Systems. 2011, pp. 2172–2179

    2011

  20. [20]

    Search-based planning for manipulation with motion primitives

    B. J. Cohen, S. Chitta, and M. Likhachev. “Search-based planning for manipulation with motion primitives”. In:IEEE International Conference on Robotics and Automation. 2010, pp. 2902–2908

    2010

  21. [21]

    Search-based motion planning for quadrotors using linear quadratic minimum time control

    S. Liu, N. Atanasov, K. Mohta, and V . Kumar. “Search-based motion planning for quadrotors using linear quadratic minimum time control”. In:IEEE/RSJ International Conference on Intelligent Robots and Systems. 2017, pp. 2872–2879

    2017

  22. [22]

    Search-Based Optimal Motion Planning for Automated Driving

    Z. Ajanovic, B. Lacevic, B. Shyrokau, M. Stolz, and M. Horn. “Search-Based Optimal Motion Planning for Automated Driving”. In: IEEE/RSJ International Conference on Intelligent Robots and Systems. 2018, pp. 4523–4530. 19

    2018

  23. [23]

    Kinodynamic RRT*: Asymptotically optimal motion planning for robots with linear dynamics

    D. J. Webb and J. van den Berg. “Kinodynamic RRT*: Asymptotically optimal motion planning for robots with linear dynamics”. In:IEEE International Conference on Robotics and Automation. 2013, pp. 5054– 5061

    2013

  24. [24]

    Optimal kinodynamic motion planning using incremental sampling-based methods

    S. Karaman and E. Frazzoli. “Optimal kinodynamic motion planning using incremental sampling-based methods”. In:IEEE Conference on Decision and Control. 2010, pp. 7681–7687

    2010

  25. [25]

    Asymptotically Optimal Planning by Feasible Kinodynamic Planning in a State–Cost Space

    K. Hauser and Y . Zhou. “Asymptotically Optimal Planning by Feasible Kinodynamic Planning in a State–Cost Space”. In:IEEE Transactions on Robotics32.6 (2016), pp. 1431–1443

    2016

  26. [26]

    Asymptotically optimal sampling-based kinodynamic planning

    Y . Li, Z. Littlefield, and K. E. Bekris. “Asymptotically optimal sampling-based kinodynamic planning”. In:The International Journal of Robotics Research35.5 (2016), pp. 528–564

    2016

  27. [27]

    KDF: Kino- dynamic Motion Planning via Geometric Sampling-Based Algorithms and Funnel Control

    C. K. Verginis, D. V . Dimarogonas, and L. E. Kavraki. “KDF: Kino- dynamic Motion Planning via Geometric Sampling-Based Algorithms and Funnel Control”. In:IEEE Transactions on Robotics39.2 (2023), pp. 978–997

    2023

  28. [28]

    Kinodynamic motion planning on roadmaps in dynamic environments

    J. van den Berg and M. Overmars. “Kinodynamic motion planning on roadmaps in dynamic environments”. In:IEEE/RSJ International Conference on Intelligent Robots and Systems. 2007, pp. 4253–4258

    2007

  29. [29]

    J. C. Butcher.Numerical Methods for Ordinary Differential Equations. John Wiley & Sons, 2016

    2016

  30. [30]

    LQR-RRT*: Optimal sampling-based motion planning with automati- cally derived extension heuristics

    A. Perez, R. Platt, G. Konidaris, L. Kaelbling, and T. Lozano-Perez. “LQR-RRT*: Optimal sampling-based motion planning with automati- cally derived extension heuristics”. In:IEEE International Conference on Robotics and Automation. 2012, pp. 2537–2542

    2012

  31. [31]

    RRT- CoLearn: towards kinodynamic planning without numerical trajectory optimization

    W. J. Wolfslag, M. Bharatheesha, T. M. Moerland, and M. Wisse. “RRT- CoLearn: towards kinodynamic planning without numerical trajectory optimization”. In:IEEE Robotics and Automation Letters3.3 (2018), pp. 1655–1662

    2018

  32. [32]

    RL-RRT: Kinodynamic motion planning via learning reachability estimators from RL policies

    H.-T. L. Chiang, J. Hsu, M. Fiser, L. Tapia, and A. Faust. “RL-RRT: Kinodynamic motion planning via learning reachability estimators from RL policies”. In:IEEE Robotics and Automation Letters4.4 (2019), pp. 4298–4305

    2019

  33. [33]

    Sampling-based kinodynamic motion planning using a neural network controller

    D. Zheng and P. Tsiotras. “Sampling-based kinodynamic motion planning using a neural network controller”. In:AIAA Scitech Forum. 2021, p. 1754

    2021

  34. [34]

    Differential flatness of mechanical control systems: A catalog of prototype systems

    R. M. Murray, M. Rathinam, and W. Sluis. “Differential flatness of mechanical control systems: A catalog of prototype systems”. In: ASME International Mechanical Engineering Congress and Exposition. 1995, pp. 349–357

    1995

  35. [35]

    Minimum snap trajectory generation and control for quadrotors

    D. Mellinger and V . Kumar. “Minimum snap trajectory generation and control for quadrotors”. In:IEEE International Conference on Robotics and Automation. 2011, pp. 2520–2525

    2011

  36. [36]

    H. K. Khalil.Nonlinear systems. Upper Saddle River, NJ: Prentice- Hall, 2002

    2002

  37. [37]

    Trajectory generation and control for precise aggressive maneuvers with quadrotors

    D. Mellinger, N. Michael, and V . Kumar. “Trajectory generation and control for precise aggressive maneuvers with quadrotors”. In:The International Journal of Robotics Research31.5 (2012), pp. 664–674

    2012

  38. [38]

    Curobo: Parallelized collision-free robot motion generation

    B. Sundaralingam, S. K. S. Hari, A. Fishman, C. Garrett, K. Van Wyk, V . Blukis, A. Millane, H. Oleynikova, A. Handa, F. Ramos, et al. “Curobo: Parallelized collision-free robot motion generation”. In:2023 IEEE International Conference on Robotics and Automation. IEEE. 2023, pp. 8112–8119

    2023

  39. [39]

    Optimal sampling-based motion planning under differential constraints: The driftless case

    E. Schmerling, L. Janson, and M. Pavone. “Optimal sampling-based motion planning under differential constraints: The driftless case”. In: IEEE International Conference on Robotics and Automation. 2015, pp. 2368–2375

    2015

  40. [40]

    Analysis of prob- abilistic roadmaps for path planning

    L. Kavraki, M. Kolountzakis, and J. -C. Latombe. “Analysis of prob- abilistic roadmaps for path planning”. In:IEEE Transactions on Robotics and Automation14.1 (1998), pp. 166–171

    1998

  41. [41]

    RRT-connect: An efficient approach to single-query path planning

    J. J. Kuffner and S. M. LaValle. “RRT-connect: An efficient approach to single-query path planning”. In:IEEE International Conference on Robotics and Automation. V ol. 2. 2000, pp. 995–1001

    2000

  42. [42]

    Kinodynamic planning

    E. Schmerling and M. Pavone. “Kinodynamic planning”. In:Encyclo- pedia of Robotics. Springer, 2019

    2019

  43. [43]

    Decoupled multiagent path planning via incremental sequential convex programming

    Y . Chen, M. Cutler, and J. P. How. “Decoupled multiagent path planning via incremental sequential convex programming”. In:IEEE International Conference on Robotics and Automation. 2015

    2015

  44. [44]

    ALTRO: A Fast Solver for Constrained Trajectory Optimization

    T. A. Howell, B. E. Jackson, and Z. Manchester. “ALTRO: A Fast Solver for Constrained Trajectory Optimization”. In:IEEE/RSJ International Conference on Intelligent Robots and Systems. 2019, pp. 7674–7679

    2019

  45. [45]

    A tutorial on Newton methods for constrained trajec- tory optimization and relations to SLAM, Gaussian Process smoothing, optimal control, and probabilistic inference

    M. Toussaint. “A tutorial on Newton methods for constrained trajec- tory optimization and relations to SLAM, Gaussian Process smoothing, optimal control, and probabilistic inference”. In:Geometric and Numerical Foundations of Movements(2017), pp. 361–392

    2017

  46. [46]

    Whole-body trajectory optimization for robot multi- modal locomotion

    G. L’Erario, G. Nava, G. Romualdi, F. Bergonti, V . Razza, S. Dafarra, and D. Pucci. “Whole-body trajectory optimization for robot multi- modal locomotion”. In:IEEE-RAS 21st International Conference on Humanoid Robots. 2022, pp. 651–658

    2022

  47. [47]

    High Accuracy Aerial Maneuvers on Legged Robots using Variational Inte- grator Discretized Trajectory Optimization

    S. Beck, C. Nguyen, T. Duong, N. Atanasov, and Q. Nguyen. “High Accuracy Aerial Maneuvers on Legged Robots using Variational Inte- grator Discretized Trajectory Optimization”. In:IEEE International Conference on Robotics and Automation. 2025, pp. 10253–10260

    2025

  48. [48]

    Motion planning around obstacles with convex optimization

    T. Marcucci, M. Petersen, D. von Wrangel, and R. Tedrake. “Motion planning around obstacles with convex optimization”. In:Science Robotics8.84 (2023)

    2023

  49. [49]

    Using Graphs of Convex Sets to Guide Nonconvex Trajectory Optimization

    D. von Wrangel and R. Tedrake. “Using Graphs of Convex Sets to Guide Nonconvex Trajectory Optimization”. In:IEEE/RSJ In- ternational Conference on Intelligent Robots and Systems. 2024, pp. 9863–9870

    2024

  50. [50]

    Towards Tight Convex Relaxations for Contact-Rich Manipulation

    B. P. Graesdal, S. Y . C. Chia, T. Marcucci, S. Morozov, A. Amice, P. Parrilo, and R. Tedrake. “Towards Tight Convex Relaxations for Contact-Rich Manipulation”. In:Robotics: Science and Systems. Delft, Netherlands, 2024

    2024

  51. [51]

    SRMP: Search-Based Robot Motion Planning Library

    I. Mishani, Y . Shaoul, R. Natarajan, J. Li, and M. Likhachev. “SRMP: Search-Based Robot Motion Planning Library”. In:arXiv preprint arXiv:2509.25352(2025)

    arXiv 2025

  52. [52]

    A formal basis for the heuristic determination of minimum cost paths

    P. E. Hart, N. J. Nilsson, and B. Raphael. “A formal basis for the heuristic determination of minimum cost paths”. In:IEEE Transac- tions on Systems Science and Cybernetics4.2 (1968), pp. 100–107

    1968

  53. [53]

    Sampling-Based Motion Planning: A Comparative Review

    A. Orthey, C. Chamzas, and L. E. Kavraki. “Sampling-Based Motion Planning: A Comparative Review”. In:Annual Review of Control, Robotics, and Autonomous Systems7.1 (July 2024), pp. 285–310

    2024

  54. [54]

    Parallel Simulation of Contact and Actuation for Soft Growing Robots

    Y . Gao, L. Chen, P. Bhovad, S. Wang, Z. Kingston, and L. H. Blumenschein. “Parallel Simulation of Contact and Actuation for Soft Growing Robots”. In:arXiv preprint arXiv:2509.15180(2025)

    arXiv 2025

  55. [55]

    Robot Motion Planning in Learned Latent Spaces

    B. Ichter and M. Pavone. “Robot Motion Planning in Learned Latent Spaces”. In:IEEE Robotics and Automation Letters4.3 (2019), pp. 2407–2414

    2019

  56. [56]

    MPC-MPNet: Model- Predictive Motion Planning Networks for Fast, Near-Optimal Planning Under Kinodynamic Constraints

    L. Li, Y . Miao, A. H. Qureshi, and M. C. Yip. “MPC-MPNet: Model- Predictive Motion Planning Networks for Fast, Near-Optimal Planning Under Kinodynamic Constraints”. In:IEEE Robotics and Automation Letters6.3 (2021), pp. 4496–4503

    2021

  57. [57]

    iDb-RRT: Sampling-based Kinodynamic Motion Planning with Motion Primitives and Trajectory Optimization

    J. Ortiz-Haro, W. H ¨onig, V . N. Hartmann, M. Toussaint, and L. Righetti. “iDb-RRT: Sampling-based Kinodynamic Motion Planning with Motion Primitives and Trajectory Optimization”. In:IEEE/RSJ International Conference on Intelligent Robots and Systems. 2024, pp. 10702–10709

    2024

  58. [58]

    PINSAT: Parallelized Interleaving of Graph Search and Trajectory Optimization for Kinodynamic Motion Planning

    R. Natarajan, S. Mukherjee, H. Choset, and M. Likhachev. “PINSAT: Parallelized Interleaving of Graph Search and Trajectory Optimization for Kinodynamic Motion Planning”. In:IEEE/RSJ International Conference on Intelligent Robots and Systems. 2024

    2024

  59. [59]

    Torque-Limited Manipulation Planning through Contact by Interleaving Graph Search and Trajectory Optimization

    R. Natarajan, G. L. Johnston, N. Simaan, M. Likhachev, and H. Choset. “Torque-Limited Manipulation Planning through Contact by Interleaving Graph Search and Trajectory Optimization”. In:IEEE International Conference on Robotics and Automation. 2023

    2023

  60. [60]

    Interleaving Graph Search and Trajectory Optimization for Aggressive Quadrotor Flight

    R. Natarajan, H. Choset, and M. Likhachev. “Interleaving Graph Search and Trajectory Optimization for Aggressive Quadrotor Flight”. In:IEEE Robotics and Automation Letters6.3 (2021), pp. 5357–5364

    2021

  61. [61]

    Sampling- based optimal kinodynamic planning with motion primitives

    B. Sakcak, L. Bascetta, G. Ferretti, and M. Prandini. “Sampling- based optimal kinodynamic planning with motion primitives”. In: Autonomous Robots43.7 (2019), pp. 1715–1732

    2019

  62. [62]

    Asymptotically optimal kinodynamic planning using bundles of edges

    R. Shome and L. E. Kavraki. “Asymptotically optimal kinodynamic planning using bundles of edges”. In:IEEE International Conference on Robotics and Automation. 2021, pp. 9988–9994

    2021

  63. [63]

    BITKOMO: Combining Sampling and Optimization for Fast Con- vergence in Optimal Motion Planning

    J. Kamat, J. Ortiz-Haro, M. Toussaint, F. T. Pokorny, and A. Orthey. “BITKOMO: Combining Sampling and Optimization for Fast Con- vergence in Optimal Motion Planning”. In:IEEE/RSJ International Conference on Intelligent Robots and Systems. 2022, pp. 4492–4497

    2022

  64. [64]

    Regionally accelerated batch informed trees (RABIT*): A framework to integrate local information into optimal path planning

    S. Choudhury, J. D. Gammell, T. D. Barfoot, S. S. Srinivasa, and S. Scherer. “Regionally accelerated batch informed trees (RABIT*): A framework to integrate local information into optimal path planning”. In:IEEE International Conference on Robotics and Automation. 2016, pp. 4207–4214

    2016

  65. [65]

    Joint sampling and trajectory optimization over graphs for online motion planning

    K. V . Alwala and M. Mukadam. “Joint sampling and trajectory optimization over graphs for online motion planning”. In:IEEE/RSJ International Conference on Intelligent Robots and Systems. 2021, pp. 4700–4707

    2021

  66. [66]

    A New Approach to Time-Optimal Path Parameterization Based on Reachability Analysis

    H. Pham and Q. -C. Pham. “A New Approach to Time-Optimal Path Parameterization Based on Reachability Analysis”. In:IEEE Transactions on Robotics34.3 (2018), pp. 645–659. 20

    2018

  67. [67]

    Jerk-limited Real-time Trajectory Generation with Arbitrary Target States

    L. Berscheid and T. Kr ¨oger. “Jerk-limited Real-time Trajectory Generation with Arbitrary Target States”. In:Robotics: Science and Systems(2021)

    2021

  68. [68]

    A real-time framework for kinodynamic planning in dynamic environments with application to quadrotor obstacle avoidance

    R. E. Allen and M. Pavone. “A real-time framework for kinodynamic planning in dynamic environments with application to quadrotor obstacle avoidance”. In:Robotics and Autonomous Systems115 (2019), pp. 174–193

    2019

  69. [69]

    Flat-RRT*: A sampling- based optimal trajectory planner for differentially flat vehicles with constrained dynamics

    L. Bascetta, I. M. Arrieta, and M. Prandini. “Flat-RRT*: A sampling- based optimal trajectory planner for differentially flat vehicles with constrained dynamics”. In:IFAC-PapersOnLine50.1 (2017), pp. 6965– 6970

    2017

  70. [70]

    Dynamically Feasible Task Space Planning for Underactuated Aerial Manipulators

    J. Welde, J. Paulos, and V . Kumar. “Dynamically Feasible Task Space Planning for Underactuated Aerial Manipulators”. In:IEEE Robotics and Automation Letters6.2 (2021), pp. 3232–3239

    2021

  71. [71]

    Efficient Sampling-based Multirotors Kinodynamic Planning with Fast Regional Optimization and Post Refining

    H. Ye, N. Pan, Q. Wang, C. Xu, and F. Gao. “Efficient Sampling-based Multirotors Kinodynamic Planning with Fast Regional Optimization and Post Refining”. In:IEEE/RSJ International Conference on Intel- ligent Robots and Systems. 2022, pp. 3356–3363

    2022

  72. [72]

    Differential Flatness-Based Trajectory Planning for Small Fixed-Wing UA Vs

    Y . Wang, W. Zeng, Y . Peng, Q. Yang, and J. Zhou. “Differential Flatness-Based Trajectory Planning for Small Fixed-Wing UA Vs”. In:International Conference on Autonomous Unmanned Systems. Springer. 2024, pp. 360–369

    2024

  73. [73]

    Exact and efficient local planning for orbitally flat systems within the RRT* framework

    M. Seemann and K. Janschek. “Exact and efficient local planning for orbitally flat systems within the RRT* framework”. In:International Conference on Control Automation Robotics & Vision. IEEE. 2014, pp. 1631–1636

    2014

  74. [74]

    An efficient spatial-temporal trajectory planner for autonomous vehicles in unstructured environments

    Z. Han, Y . Wu, T. Li, L. Zhang, L. Pei, L. Xu, C. Li, C. Ma, C. Xu, S. Shen, et al. “An efficient spatial-temporal trajectory planner for autonomous vehicles in unstructured environments”. In: IEEE Transactions on Intelligent Transportation Systems25.2 (2023), pp. 1797–1814

    2023

  75. [75]

    Differential flatness-based trajectory planning for multiple unmanned aerial vehicles using mixed- integer linear programming

    Y . Hao, A. Davari, and A. Manesh. “Differential flatness-based trajectory planning for multiple unmanned aerial vehicles using mixed- integer linear programming”. In:American Control Conference. 2005, pp. 104–109

    2005

  76. [76]

    Optimal control of differen- tially flat systems is surprisingly easy

    L. E. Beaver and A. A. Malikopoulos. “Optimal control of differen- tially flat systems is surprisingly easy”. In:Automatica159 (2024), p. 111404

    2024

  77. [77]

    Sampling-based trajectory (re) planning for differentially flat systems: Application to a 3D gantry crane

    M. N. Vu, M. Schwegel, C. Hartl-Nesic, and A. Kugi. “Sampling-based trajectory (re) planning for differentially flat systems: Application to a 3D gantry crane”. In:IFAC-PapersOnLine55.38 (2022), pp. 33–40

    2022

  78. [78]

    A little more, a lot better: Improving path quality by a path-merging algorithm

    B. Raveh, A. Enosh, and D. Halperin. “A little more, a lot better: Improving path quality by a path-merging algorithm”. In:IEEE Transactions on Robotics27.2 (2011), pp. 365–371

    2011

  79. [79]

    Probabilistic roadmap methods are embarrassingly parallel

    N. M. Amato and L. K. Dale. “Probabilistic roadmap methods are embarrassingly parallel”. In:IEEE International Conference on Robotics and Automation. V ol. 1. 1999, pp. 688–694

    1999

  80. [80]

    Parallel sampling-based motion planning with superlinear speedup

    J. Ichnowski and R. Alterovitz. “Parallel sampling-based motion planning with superlinear speedup”. In:IEEE/RSJ International Conference on Intelligent Robots and Systems. IEEE. 2012, pp. 1206– 1212

    2012

Showing first 80 references.