Pith. sign in

REVIEW 3 major objections 5 minor 1 cited by

Decentralized Uncertainty-Aware Multi-Agent Collision Avoidance with Model Predictive Path Integral

T0 review · 3 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read A decentralized MPPI controller with chance-constrained ORCA safety buffers achieves high success rates for differential-drive robots under observation and execution noise.

desk verdict Useful practical extension of MPPI-ORCA to sensing and execution noise, with strong simulation results, but the stated probability guarantee does not actually apply to the control that MPPI executes. read the letter →

arxiv 2507.20293 v2 pith:WF2DUFHF submitted 2025-07-27 cs.RO

Stepan Dergachev , Konstantin Yakovlev This is my paper · ORCID
classification cs.RO
keywords approachavoidancecollisionmulti-agentconstraintsdecentralizedintegralmodel
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

Robots navigating among each other usually assume they know exactly where the others are and that their own movement commands execute perfectly. In reality, sensors give noisy positions and velocities, and motors do not exactly follow commands. This paper addresses that gap for a popular control method called Model Predictive Path Integral (MPPI), which works by sampling many possible control sequences and picking the best one.

The authors add safety buffers to the collision-avoidance constraints. For observation noise, they inflate the other robot's radius using a chi-square bound from the noise covariance. For execution noise, they turn each safety constraint into a chance constraint, a form that guarantees the constraint holds with a chosen probability, and they solve a convex optimization problem to adjust the sampling distribution. This way, the sampled controls are more likely to be safe.

They test with 2 to 25 differential-drive robots in two standard scenarios, with injected Gaussian noise, and compare against three baselines. Their method reaches 100% success in all tested instances, while baselines often collide or stall. They also validate in the Gazebo simulator with ten TurtleBot robots. However, the theoretical safety guarantee is only for individual sampled controls, not explicitly for the final control that MPPI computes as a weighted average, and the joint safety probability across multiple neighbors is not rigorously bounded.

Extended reading notes

Core claim

By solving this optimization problem, we obtain new parameters for the sampling distribution, ensuring that with probability at least δc = δo × δν × δu, the sampled controls will be safe. (Section V-B, after Eq. 24). If correct, the method provides probabilistic collision avoidance for differential-drive robots under known Gaussian sensing and execution noise, and outperforms ORCA-DD, B-UAVC, and MPPI-ORCA in the tested scenarios.

Load-bearing premise

The paper assumes that making each sampled control safe with probability δc implies the final MPPI control, which is a weighted average over samples, is also safe with at least that probability. This implication is not proven, and the claimed joint probability does not account for multiple neighboring agents or dependence between constraints. If this fails, the stated safety guarantee does not hold for the executed control. (Section V-B, Eq. (24) and following paragraph).

Share X Bluesky LinkedIn Reddit HN

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. This paper proposes a decentralized multi-agent collision avoidance method that combines Model Predictive Path Integral (MPPI) with probabilistic ORCA-style linear constraints. At each control step, the agent solves a Second-Order Cone Program to adjust the mean and covariance of the MPPI sampling distribution so that sampled controls satisfy chance constraints accounting for observation noise and execution noise. The authors validate the method in numerical simulations of differential-drive robots against ORCA-DD, B-UAVC, and MPPI-ORCA, and in a Gazebo simulation with TurtleBot 3 robots, reporting 100% success rates in the tested scenarios and providing a public source code repository.

Significance. The integration of MPPI with uncertainty-aware chance constraints is timely and practically relevant. The per-sample chance-constraint conversion is standard, and the SOCP formulation is a clean extension of the authors' prior MPPI-ORCA work. The empirical evaluation is thorough, with multiple scenarios, a realistic Gazebo validation, and comparisons to three baselines; the availability of source code is a strength. However, the paper's central formal claim—that the executed control is safe with probability δc—is not supported by the presented derivation, which only addresses individual samples. If that gap is fixed, the contribution would be significant for the community.

major comments (3)
  1. [Section V-B, Eq. (24), and Section IV-A, Eq. (9)] The safety guarantee stated after Eq. (24) applies to sampled control sequences, but the control actually executed by MPPI is the weighted average u* = Σ_k ω(U^k) u^k from Eq. (9). The paper does not prove that this weighted average inherits the per-sample chance constraint. Since the safe set defined by the linear ORCA constraints is convex, a sufficient condition for u* to be safe is that every sample is safe, which occurs with probability at most δc^K (or lower if weights are correlated with constraint violation). For typical K on the order of hundreds or thousands and δc ≈ 0.995, this probability is negligibly small. Consequently, the problem statement's requirement (iii), that the control input be probabilistically safe, is not established for the executed control. The authors should either provide a formal argument for the weighted-average control or explicitly restrict the claim to the samples and adjust the abstract and conclusions accordingly.
  2. [Section V-B, Eqs. (20)-(24)] The joint probability δc = δo × δν × δu is presented as the probability that a sampled control is safe, but the optimization problem (24) contains one chance constraint per neighbor j ∈ A'_i. The product formula does not combine multiple constraints: for a sample to be safe with respect to all observed robots, all constraints must hold simultaneously. By the union bound, the probability of joint satisfaction is at least 1 - n(1 - δu δν δo) for n neighbors, which is substantially lower than the claimed product when n > 1 and the per-constraint confidence is high. The same union-bound issue applies to the observation confidence δo, which is per observed neighbor. The paper should state the correct n-dependent bound or use a joint chance constraint, and the empirical parameters (δu=δν=0.999, δo=0.9975) should be presented in terms of the resulting per-sample, all-neighbor guarantee.
  3. [Section V-B, Eq. (24) and Section VI-A-1] The relationship between the claimed guarantee and the implemented controller needs clarification. If the algorithm executes the optimized mean μ' rather than the MPPI weighted average of Eq. (9), then the stochasticity of the command is not present in the executed control, and the δu factor in Eq. (22) is spurious; the safety probability would be at most δo δν. If the algorithm does execute the weighted average, the first major comment applies. The manuscript should specify which quantity is commanded and tailor the probability statement to that quantity.
minor comments (5)
  1. [Section III] The threshold δ in the definition of a probabilistically safe control is never defined; the later text introduces δo, δν, and δu but does not relate them to δ.
  2. [Section VI-A-1] The sentence 'For ORCA-MPPI and our method, the probability of sampling a control action outside the safety constraints, denoted as δu, was set to 0.999' appears to describe the confidence level (probability of staying inside) but is worded as a probability of violation; please correct the wording.
  3. [Section V-A] The derivation of the inflated radius ro uses χ^2_2^{-1}(δo), but the text does not justify this choice for the case of correlated position estimates; a brief explanation or reference would help.
  4. [Equation (24)] There is a formatting error in the left-hand side of the first constraint ('a′jΣ′a′Tj' should be 'a′_j^T Σ′ a′_j'); also, the objective's norm notation is ambiguous about which vector norm is used in the experiments.
  5. [Section V-C] The cost function introduces d_th and d_la and a Kalman filter without specifying the filter parameters or how the predicted positions are obtained; please provide details or a reference.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the safety derivation is a chance-constrained optimization with user-specified confidence levels; the only self-citation ([10]) is a baseline/implementation reference, not a load-bearing proof.

full rationale

The paper's central derivation (Section V-B, Eqs. (20)-(24)) constructs safety constraints from user-chosen confidence levels δo, δν, δu and standard Gaussian/chi-square quantiles. The SOCP objective (15) minimizes deviation from the original sampling parameters while enforcing these chance constraints; nothing is fitted to the experimental outcomes and then reported as a prediction. The claim that a sampled control is safe with probability at least δoδνδu is a composition of the three chance bounds, not a restatement of the conclusion in the premises: the δ values are inputs, and the safety probability is a derived bound. The asserted transfer from per-sample safety to the weighted-average MPPI control of Eq. (9) is not proven, and the product bound ignores multiple neighbors; however, that is a correctness/derivation gap, not circularity, because the executed average is not defined in terms of the claimed safety probability. The repeated citation of the authors' prior MPPI-ORCA [10] supplies the baseline algorithm and the SOCP implementation style, but the present paper states the full optimization problem and validates against external benchmarks (ORCA-DD, B-UAVC, Nav2/Gazebo), so the self-citation is not load-bearing in the sense of making the central claim true by construction. Under the rubric this is a minor, non-load-bearing self-citation at most, hence score 2.

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

The central claims rest mainly on known probability facts and standard control-theoretic assumptions. The main extra burden is the user-chosen confidence levels and cost weights, which are not given explicitly, and the unproven transfer of individual sample safety to the weighted-average control.

free parameters (4)
  • δo, δν, δu = 0.9975, 0.999, 0.999
    User-selected confidence levels that define the claimed safety probability; they directly determine the buffer sizes and the stated guarantee.
  • Cost weights w_term, w_goal, w_dist, w_col, w_vel = Not reported in the paper.
    Hand-tuned weights in the MPPI cost function; values are not given in the paper, affecting reproduction.
  • Look-ahead distance d_la and distance threshold d_th = Not reported in the paper.
    Parameters of the goal progress and distance costs; not specified in the text.
  • Sampling variance multiplier K = Not reported in the paper.
    MPPI sampling covariance Σ* = K Σ; the paper mentions it but does not give the value used.
assumptions (3)
  • domain assumption Observation noise and execution noise are zero-mean Gaussian with known covariances Σ_p, Σ_v, Σ.
    The method relies on these distributions to compute buffers and chance constraints (Section III, Eq. 1, and the observation model).
  • domain assumption All agents are homogeneous and follow the same avoidance policy, enabling reciprocal assumptions.
    ORCA's reciprocal avoidance assumption is carried over; the paper states homogeneity in Section III.
  • standard math The true position of a neighbor lies within the chi-square confidence ellipse with probability δo.
    Uses χ² quantile with 2 degrees of freedom and the largest eigenvalue of Σ_p (Eq. 14).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Decentralized Uncertainty-Aware Multi-Agent Collision Avoidance with Model Predictive Path Integral." pith.science (2026). https://pith.science/paper/WF2DUFHF

@misc{pith2026250720293,
  author       = {Pith},
  title        = {Pith review of: Decentralized Uncertainty-Aware Multi-Agent Collision Avoidance with Model Predictive Path Integral},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WF2DUFHF}},
  note         = {Machine review of arXiv:2507.20293}
}
read the original abstract

Decentralized multi-agent navigation under uncertainty is a complex task that arises in numerous robotic applications. It requires collision avoidance strategies that account for both kinematic constraints, sensing and action execution noise. In this paper, we propose a novel approach that integrates the Model Predictive Path Integral (MPPI) with a probabilistic adaptation of Optimal Reciprocal Collision Avoidance. Our method ensures safe and efficient multi-agent navigation by incorporating probabilistic safety constraints directly into the MPPI sampling process via a Second-Order Cone Programming formulation. This approach enables agents to operate independently using local noisy observations while maintaining safety guarantees. We validate our algorithm through extensive simulations with differential-drive robots and benchmark it against state-of-the-art methods, including ORCA-DD and B-UAVC. Results demonstrate that our approach outperforms them while achieving high success rates, even in densely populated environments. Additionally, validation in the Gazebo simulator confirms its practical applicability to robotic platforms. A source code is available at http://github.com/PathPlanning/MPPI-Collision-Avoidance.

Figures

Figures reproduced from arXiv: 2507.20293 by the authors.

Figure 1
Figure 1. The decentralized multi-agent collision avoidance [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. An example of safe control sampling (velocity sam [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. The average makespan for evaluated algorithms for different number of agents in Circle and Random scenarios. Each point on the graph is included only if more than 50% of the runs were successful. The lower is the better. positions and velocities of the neighboring agents were supplemented with noise with zero mean and covariance Σp = Σv = diag(0.1,0.1) 2 . Two types of scenarios, widely used in literature on colli￾s… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Simulation in Gazebo. (a) 10 Turtlebots executing the mission, (b) MPPI rollouts with sampled trajectories, (c)-(d) [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

  1. CoRL-MPPI: Enhancing MPPI With Learnable Behaviours For Efficient And Provably-Safe Multi-Robot Collision Avoidance

    cs.RO 2025-11 conditional novelty 5.0 of 10

    CoRL-MPPI injects a learned cooperative policy into MPPI's sampling distribution to speed up and make safer multi-robot navigation in dense simulations.

Reference graph

Works this paper leans on

36 extracted references · 31 canonical work pages · cited by 1 Pith paper

  1. [1]

    Reciprocal n-body collision avoidance,

    J. Van Den Berg, S. J. Guy, M. Lin, and D. Manocha, “Reciprocal n-body collision avoidance,” in Robotics research, 2011, pp. 3–19

    work page 2011

  2. [2]

    Fast, on- line collision avoidance for dynamic vehicles using buffered voronoi cells,

    D. Zhou, Z. Wang, S. Bandyopadhyay, and M. Schwager, “Fast, on- line collision avoidance for dynamic vehicles using buffered voronoi cells,” IEEE Robotics and Automation Letters , vol. 2, no. 2, pp. 1047– 1054, 2017

    work page 2017

  3. [3]

    The hybrid reciprocal velocity obstacle,

    J. Snape, J. Van Den Berg, S. J. Guy, and D. Manocha, “The hybrid reciprocal velocity obstacle,” IEEE Transactions on Robotics , vol. 27, no. 4, pp. 696–706, 2011

    work page 2011

  4. [4]

    Reciprocal collision avoidance with acceleration-velocity obstacles,

    J. Van Den Berg, J. Snape, S. J. Guy, and D. Manocha, “Reciprocal collision avoidance with acceleration-velocity obstacles,” in 2011 IEEE International Conference on Robotics and Automation . IEEE, 2011, pp. 3475–3482

    work page 2011

  5. [5]

    Optimal reciprocal collision avoidance for multiple non- holonomic robots,

    J. Alonso-Mora, A. Breitenmoser, M. Rufli, P. Beardsley, and R. Sieg- wart, “Optimal reciprocal collision avoidance for multiple non- holonomic robots,” in Distributed autonomous robotic systems: The 10th international symposium , 2013, pp. 203–216

    work page 2013

  6. [6]

    Decentralized probabilis- tic multi-robot collision avoidance using buffered uncertainty-aware voronoi cells,

    H. Zhu, B. Brito, and J. Alonso-Mora, “Decentralized probabilis- tic multi-robot collision avoidance using buffered uncertainty-aware voronoi cells,” Autonomous Robots, vol. 46, no. 2, pp. 401–420, 2022

    work page 2022

  7. [7]

    Collision avoidance in pedestrian-rich environments with deep reinforcement learning,

    M. Everett, Y . F. Chen, and J. P. How, “Collision avoidance in pedestrian-rich environments with deep reinforcement learning,” IEEE Access, vol. 9, pp. 10 357–10 377, 2021

    work page 2021

  8. [8]

    Aggressive driving with model predictive path integral control,

    G. Williams, P. Drews, B. Goldfain, J. M. Rehg, and E. A. Theodorou, “Aggressive driving with model predictive path integral control,” in 2016 IEEE International Conference on Robotics and Automation (ICRA), 2016, pp. 1433–1440

    2016

Show all 36 references
  1. [9]

    Information theoretic mpc for model-based reinforcement learning,

    G. Williams, N. Wagener, B. Goldfain, P. Drews, J. M. Rehg, B. Boots, and E. A. Theodorou, “Information theoretic mpc for model-based reinforcement learning,” in 2017 IEEE International Conference on Robotics and Automation (ICRA) , 2017, pp. 1714–1721

    2017

  2. [10]

    Model predictive path integral for de- centralized multi-agent collision avoidance,

    S. Dergachev and K. Yakovlev, “Model predictive path integral for de- centralized multi-agent collision avoidance,” PeerJ Computer Science, vol. 10, p. e2220, 2024

    2024

  3. [11]

    Smooth and collision-free navigation for multiple robots under differential-drive constraints,

    J. Snape, J. Van Den Berg, S. J. Guy, and D. Manocha, “Smooth and collision-free navigation for multiple robots under differential-drive constraints,” in 2010 IEEE/RSJ international conference on intelligent robots and systems , 2010, pp. 4584–4589

    2010

  4. [12]

    Chance-constrained collision avoidance for mavs in dynamic environments,

    H. Zhu and J. Alonso-Mora, “Chance-constrained collision avoidance for mavs in dynamic environments,” IEEE Robotics and Automation Letters, vol. 4, no. 2, pp. 776–783, 2019

    2019

  5. [13]

    Smooth coordination and navigation for multiple differential-drive robots,

    J. Snape, S. J. Guy, J. Van Den Berg, and D. Manocha, “Smooth coordination and navigation for multiple differential-drive robots,” in Experimental Robotics: The 12th International Symposium on Experimental Robotics , 2014, pp. 601–613

    2014

  6. [14]

    Cooperative collision avoidance for nonholonomic robots,

    J. Alonso-Mora, P. Beardsley, and R. Siegwart, “Cooperative collision avoidance for nonholonomic robots,” IEEE Transactions on Robotics , vol. 34, no. 2, pp. 404–420, 2018

    2018

  7. [15]

    Prvo: Probabilistic reciprocal velocity obstacle for multi robot navigation under uncertainty,

    B. Gopalakrishnan, A. K. Singh, M. Kaushik, K. M. Krishna, and D. Manocha, “Prvo: Probabilistic reciprocal velocity obstacle for multi robot navigation under uncertainty,” in 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , 2017, pp. 1089– 1096

    2017

  8. [16]

    Multi-robot col- lision avoidance with localization uncertainty,

    D. Hennes, D. Claes, W. Meeussen, and K. Tuyls, “Multi-robot col- lision avoidance with localization uncertainty,” in Proceedings of the 11th International Conference on Autonomous Agents and Multiagent Systems-V olume 1, 2012, pp. 147–154

    2012

  9. [17]

    Collision avoidance under bounded localization uncertainty,

    D. Claes, D. Hennes, K. Tuyls, and W. Meeussen, “Collision avoidance under bounded localization uncertainty,” in 2012 IEEE/RSJ Interna- tional Conference on Intelligent Robots and Systems , 2012, pp. 1192– 1198

    2012

  10. [18]

    Distributed collision avoidance of multiple robots with probabilistic buffered voronoi cells,

    M. Wang and M. Schwager, “Distributed collision avoidance of multiple robots with probabilistic buffered voronoi cells,” in 2019 in- ternational symposium on multi-robot and multi-agent systems (MRS) , 2019, pp. 169–175

    2019

  11. [19]

    Deep-learned collision avoidance policy for distributed multiagent navigation,

    P. Long, W. Liu, and J. Pan, “Deep-learned collision avoidance policy for distributed multiagent navigation,” IEEE Robotics and Automation Letters, vol. 2, no. 2, pp. 656–663, 2017

    2017

  12. [20]

    To- wards optimally decentralized multi-robot collision avoidance via deep reinforcement learning,

    P. Long, T. Fan, X. Liao, W. Liu, H. Zhang, and J. Pan, “To- wards optimally decentralized multi-robot collision avoidance via deep reinforcement learning,” in 2018 IEEE international conference on robotics and automation (ICRA) , 2018, pp. 6252–6259

    2018

  13. [21]

    Crowd-robot interaction: Crowd-aware robot navigation with attention-based deep reinforce- ment learning,

    C. Chen, Y . Liu, S. Kreiss, and A. Alahi, “Crowd-robot interaction: Crowd-aware robot navigation with attention-based deep reinforce- ment learning,” in 2019 international conference on robotics and automation (ICRA) , 2019, pp. 6015–6022

    2019

  14. [22]

    Distributed multi-robot collision avoidance via deep reinforcement learning for navigation in complex scenarios,

    T. Fan, P. Long, W. Liu, and J. Pan, “Distributed multi-robot collision avoidance via deep reinforcement learning for navigation in complex scenarios,” The International Journal of Robotics Research , vol. 39, no. 7, pp. 856–892, 2020

    2020

  15. [23]

    Robust model predictive path integral control: Analysis and performance guarantees,

    M. S. Gandhi, B. Vlahov, J. Gibson, G. Williams, and E. A. Theodorou, “Robust model predictive path integral control: Analysis and performance guarantees,” IEEE Robotics and Automation Letters , vol. 6, no. 2, pp. 1423–1430, 2021

    2021

  16. [24]

    Constrained covariance steering based tube-mppi,

    I. M. Balci, E. Bakolas, B. Vlahov, and E. A. Theodorou, “Constrained covariance steering based tube-mppi,” in 2022 American Control Conference (ACC), 2022, pp. 4197–4202

    2022

  17. [25]

    Control barrier function augmentation in sampling-based control algorithm for sample efficiency,

    C. Tao, H. Kim, H. Yoon, N. Hovakimyan, and P. V oulgaris, “Control barrier function augmentation in sampling-based control algorithm for sample efficiency,” in 2022 American Control Conference (ACC) , 2022, pp. 3488–3493

    2022

  18. [26]

    Path integral methods with stochastic control barrier functions,

    C. Tao, H.-J. Yoon, H. Kim, N. Hovakimyan, and P. V oulgaris, “Path integral methods with stochastic control barrier functions,” in 2022 IEEE 61st Conference on Decision and Control (CDC) , 2022, pp. 1654–1659

    2022

  19. [27]

    Smooth model predictive path integral control without smoothing,

    T. Kim, G. Park, K. Kwak, J. Bae, and W. Lee, “Smooth model predictive path integral control without smoothing,” IEEE Robotics and Automation Letters , vol. 7, no. 4, pp. 10 406–10 413, 2022

    2022

  20. [28]

    Towards efficient mppi trajectory generation with unscented guidance: U-mppi control strategy,

    I. S. Mohamed, J. Xu, G. S. Sukhatme, and L. Liu, “Towards efficient mppi trajectory generation with unscented guidance: U-mppi control strategy,” IEEE Transactions on Robotics , 2025

    2025

  21. [29]

    Sampling-based optimization for multi-agent model predictive con- trol,

    Z. Wang, A. D. Saravanos, H. Almubarak, O. So, and E. A. Theodorou, “Sampling-based optimization for multi-agent model predictive con- trol,” arXiv preprint arXiv:2211.11878 , 2022

    2022 arXiv

  22. [30]

    Safety embedded stochastic optimal control of networked multi-agent systems via barrier states,

    L. Song, P. Zhao, N. Wan, and N. Hovakimyan, “Safety embedded stochastic optimal control of networked multi-agent systems via barrier states,” in 2023 American Control Conference (ACC) , 2023, pp. 2554– 2559

    2023

  23. [31]

    Chance-constrained sampling- based mpc for collision avoidance in uncertain dynamic environ- ments,

    I. S. Mohamed, M. Ali, and L. Liu, “Chance-constrained sampling- based mpc for collision avoidance in uncertain dynamic environ- ments,” arXiv preprint arXiv:2501.08520 , 2025

    2025 arXiv

  24. [32]

    Multi-agent path integral control for interaction-aware motion planning in urban canals,

    L. Streichenberg, E. Trevisan, J. J. Chung, R. Siegwart, and J. Alonso- Mora, “Multi-agent path integral control for interaction-aware motion planning in urban canals,” in 2023 IEEE International Conference on Robotics and Automation (ICRA) , 2023, pp. 1379–1385

    2023

  25. [33]

    Motion planning in dynamic environments using velocity obstacles,

    P. Fiorini and Z. Shiller, “Motion planning in dynamic environments using velocity obstacles,” The international journal of robotics re- search, vol. 17, no. 7, pp. 760–772, 1998

    1998

  26. [34]

    Probabilistic collision checking with chance constraints,

    N. E. Du Toit and J. W. Burdick, “Probabilistic collision checking with chance constraints,” IEEE Transactions on Robotics, vol. 27, no. 4, pp. 809–815, 2011

    2011

  27. [35]

    The marathon 2: A navigation system,

    S. Macenski, F. Martin, R. White, and J. Gin ´es Clavero, “The marathon 2: A navigation system,” in 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , 2020

    2020

  28. [36]

    The dynamic window approach to collision avoidance,

    D. Fox, W. Burgard, and S. Thrun, “The dynamic window approach to collision avoidance,” IEEE Robotics & Automation Magazine , vol. 4, no. 1, pp. 23–33, 1997

    1997

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.