Pith. sign in

REVIEW 4 major objections 5 minor 80 references

Swept Volume-Aware Trajectory Planning and MPC Tracking for Multi-Axle Swerve-Drive AMRs

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read A swept-area-aware SDF planner plus MPC tracking cuts a five-axle AMR's excess swept area from 48.37 m² to 23.14 m² in a simulated left turn.

desk verdict Real problem, plausible pipeline, but the MPC plant model and wheel-velocity decomposition use the wrong kinematics, so the validation does not yet transfer to hardware. read the letter →

arxiv 2412.16875 v2 pith:6SJ2ZTQJ submitted 2024-12-22 cs.RO

Tianxin Hu , Shenghai Yuan , Ruofei Bai , Xinghang Xu , Yuwen Liao , Fen Liu , Lihua Xie This is my paper · ORCID
classification cs.RO
keywords sweptvolumeminimizationmulti-axleAMRswerve-drivemodelpredictivecontrolsigneddistancefieldpathplanningtrajectoryoptimizationindependentwheelsteeringautonomouslogistics
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper is trying to show that the swept volume, the extra ground a long multi-axle vehicle occupies during a turn, can be actively minimized by coupling a swept-area-aware path planner with a model predictive controller that steers each wheel independently. The proposed pipeline plans a path with obstacle avoidance and a cost that keeps the vehicle's length axis aligned with its direction of travel, then tracks that path with MPC and converts the body-level velocity command into per-wheel steering angles. In a simulated left-turn scenario with a five-axle, 8.1-meter vehicle, the method reports an excess swept area of 23.14 m², about half the 48.37 m² of the best comparison method, along with 1.17 seconds of planning time and lateral and heading tracking errors within ±0.04 m and ±0.03°. If these results hold, warehouse and highway logistics vehicles could maneuver through tight spaces with smaller safety zones, reducing the risk of hitting pedestrians or obstacles. The paper also claims this is the first comprehensive approach to combine swept-volume minimization with independent axle control.

What carries the argument

The load-bearing object is the swept-area cost Jsv = Σ (Δφ_j)², with Δφ_j = φ_j − atan(VY,j/VX,j). This is a heading-velocity alignment penalty: at each control point it drives the vehicle's heading φ_j to coincide with the direction of its velocity vector, so the long axis of the rectangular body stays tangent to the path and the vehicle sweeps a narrower corridor. It sits inside the second optimization stage alongside an SDF-based obstacle cost Job that penalizes proximity to obstacles using the implicit signed distance field of the rectangular footprint, and the whole trajectory is parameterized as MINCO, the minimum-control-effort polynomial trajectory class, so gradients with respect to control points and segment times are available. The MPC layer then takes the optimized body-level control (Vx, Vy, ω) and derives each wheel's steering angle γ_i and speed Vi through the rigid-body velocity relation, which is what turns a point-mass optimal control problem into commands for independently steerable axles.

What would settle it

In the same Gazebo left-turn scene, rasterize the vehicle footprint at every pose along the planned and tracked trajectory, take the union to get the actual swept area, and compute S_excess; then rerun the second optimization stage with W_sv = 0. If removing the swept-area cost does not increase the measured excess area, or if trajectories with lower J_sv do not have smaller S_excess, the claimed mechanism is not what produces the reported reduction.

Watch

Extended reading notes

Core claim

The central claim is that swept-volume minimization for multi-axle swerve-drive AMRs can be achieved by a two-stage optimization followed by MPC tracking: first smooth an A* path, then optimize it under SDF obstacle costs and a swept-area cost Jsv that penalizes the squared angular difference between the vehicle heading φ_j and the velocity direction atan(VY,j/VX,j), and finally track it with an MPC controller that converts the optimal body velocity (Vx, Vy, ω) into per-wheel steering angles γ_i = arctan(Viy/Vix). In their Gazebo left-turn scenario, this reduces excess swept area to 23.14 m² versus 48.37 m² for the hardest baseline, with planning time of 1.17 seconds and tracking errors within ±0.04 m and ±0.03°. The paper presents this as the first comprehensive approach to combine these elements.

Load-bearing premise

The entire swept-volume reduction rests on the assumption that penalizing the squared angle between the vehicle's heading and its velocity direction is a faithful proxy for minimizing the actual swept area, since the optimizer never directly computes or evaluates that area.

Editorial extensions

If this is right

  • In the simulated left-turn scenario, the proposed pipeline reduces excess swept area from 48.37 m² for the best baseline to 23.14 m², roughly halving the extra ground covered.
  • Planning time of 1.17 seconds with CUDA-accelerated swept-volume SDF estimation is fast enough for near-real-time replanning in a static scenario.
  • The MPC tracker holds lateral error within ±0.04 m and heading error within ±0.03° on the planned trajectory, so the theoretical swept-area reduction is not lost to tracking deviations.
  • Computing per-wheel steering angles from the velocity vector allows a single point-mass MPC controller to command all independently steerable wheels, removing the need for a separate steering-center mode switch.
  • SVSDF-style planning can be combined with this MPC tracker and still achieve accurate tracking, indicating the tracking component generalizes beyond the paper's own planner.

Reading between the lines

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

  • For a fixed chassis, the benefit should scale with vehicle length and the number of steerable axles; a short single-axle robot would likely show little difference, so the method is most relevant to long multi-axle platforms.
  • The heading-alignment idea could transfer to articulated vehicles by imposing a per-trailer alignment cost, though the paper does not test this.
  • Because the MPC tracker is separable from the planner, the same per-wheel steering conversion could be applied to any path planner, not just the SDF-based one used here.
  • Pairing the swept-area proxy with a swept-region SDF, so that the planner checks collisions against the area the vehicle will actually occupy, could remove the need to trust the proxy; this would be a natural follow-up.
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

4 major / 5 minor

Summary. The paper proposes a unified framework for multi-axle swerve-drive automated mobile robots that combines signed-distance-field (SDF) based trajectory planning with model predictive control (MPC) and per-wheel steering-angle conversion, with the goal of minimizing swept volume during turns. The planner first generates an A* path, smooths it with MINCO polynomials, and then optimizes a cost that includes obstacle distance, energy, time, and a heading-velocity alignment term intended to reduce swept area. The tracking layer uses a linear MPC with state X=[x,y,φ]^T and control u=[Vx,Vy,ω]^T, followed by a conversion to individual wheel speeds and steering angles. The method is evaluated in a Gazebo simulation of a 5-axle vehicle making a left turn, reporting an excess swept area of 23.14 m² versus 48.37 m² for the closest baseline, along with small tracking errors. The paper claims this is the first comprehensive approach to combine swept-volume minimization with independent axle control.

Significance. If the kinematic models and swept-area objective were correct, the paper would address a practically important problem for large multi-axle vehicles in constrained logistics environments, and the use of CUDA-accelerated swept-volume SDF evaluation plus an MPC tracking layer is a plausible pipeline. The authors also state an intent to open-source the work, which is commendable. However, the significance is currently undermined by load-bearing errors in the vehicle kinematics used for both MPC prediction and wheel command generation, and by the lack of a direct relationship between the optimized heading-velocity cost and the actual swept area. These issues mean the reported simulation improvements are not trustworthy evidence that the method would work on a real vehicle.

major comments (4)
  1. [III.C, Eq. (22)] The discrete-time plant model X(k+1)=A X(k)+B u(k) with A=I, B=T·I, and u=[Vx,Vy,ω]^T is not a valid kinematic model for a rigid body. If Vx and Vy are body-frame velocities (as stated in Section III.A, where they are called longitudinal and lateral velocities), then the position update must depend on heading: x(k+1)=x(k)+T(Vx cosφ − Vy sinφ), y(k+1)=y(k)+T(Vx sinφ + Vy cosφ), φ(k+1)=φ(k)+Tω. If instead Vx and Vy are global-frame velocities, then the wheel decomposition in Eq. (28) is inconsistent because it adds body-frame rotational terms ωYwi and ωXwi to global components. Under either interpretation the MPC prediction does not describe the actual kinematics, so the small tracking errors in Table I could be artifacts of a simulator using the same incorrect model rather than evidence of real tracking performance.
  2. [III.C, Eq. (28)] The wheel-velocity decomposition contains a sign error in the transverse component. For a point fixed in the body frame at coordinates (Xwi, Ywi), with counterclockwise-positive yaw rate ω, the rigid-body velocity is Vix = Vx − ωYwi and Viy = Vy + ωXwi. Equation (28) instead gives Vix = Vx + ωYwi and Viy = Vy + ωXwi. This changes the computed wheel steering angles γi and speeds Vi in Eqs. (29)–(30), and if the same erroneous kinematics are implemented in the Gazebo simulation, the reported trajectory tracking and swept-area results do not demonstrate correct behavior on a physical swerve-drive AMR.
  3. [III.B.2, Eqs. (15)–(16)] The swept-area cost Jsv is a heading-velocity alignment penalty, not the swept area S. The paper asserts that aligning the vehicle's long axis with the instantaneous velocity direction reduces swept area, but this is not established for a multi-axle vehicle whose rear axles off-track whenever yaw is nonzero. The optimizer never evaluates S during planning, so the claim that the planner 'minimizes swept volume' is not supported. A direct demonstration is needed, for example an ablation that compares the proposed proxy against an optimization that actually uses the swept-area SDF, or an analysis showing that the proxy bounds the true swept area. The post-hoc Sexcess metric in Section IV-C cannot substitute for a validated optimization objective.
  4. [IV-C, Table I] The experimental comparison conflates the planner and the controller. The Classic and Hierarchical baselines track the trajectory generated by the proposed planner but do not receive the swept-volume objective, so the Sexcess improvement in Table I may come primarily from the MPC tracking layer rather than from swept-volume-aware planning. The SVSDF baseline uses the proposed MPC but not the swept-area cost. To support the central claim, the comparison needs baselines that isolate the contribution of the Jsv term with a correct kinematic model, and the 'minimal swept area' used to define Sexcess must be precisely specified (for example, the convex hull of the vehicle's footprint, or the area of a straight corridor of vehicle width along the reference path).
minor comments (5)
  1. [III.B.1, Eq. (4)] The expression ∂Pj/∂Tj = Vj is dimensionally inconsistent as written: Pj is a position vector and Vj is a velocity, so the equality holds only under a specific time-scaling convention for the MINCO trajectory that should be stated explicitly.
  2. [III.B.2, Eq. (20)] The chain rule in Eq. (20) relies on the same derivative ∂Pj/∂Tj = Vj as Eq. (4), and therefore inherits the same need for a clearly defined time-scaling convention; as written, the terms ∂Jsv/∂Xj·VX,j and ∂Jsv/∂Yj·VY,j are also mixing position and velocity coordinates.
  3. [IV-C, Table I and Fig. 6] The tracking errors ey and eφ are reported only as ranges (e.g., ±0.04 m, ±0.03°), without the number of trials, the duration of each run, or the statistical variation, so it is unclear whether these represent worst-case, 1σ, or peak values over a single trajectory.
  4. [I, Introduction] The statement 'A critical challenges is the swept volume' contains a grammatical error, and the terms 'swept volume' and 'swept area' are used interchangeably earlier in the text before the constant-height equivalence is introduced in Section III.A; this should be made consistent.
  5. [Abstract and V, Conclusion] The claim that the approach 'delivers life-saving improvements' is too strong for a simulation-only validation with an unverified kinematic model, and the promise of an open-source release should be accompanied by a functional repository link or a code availability statement in the manuscript.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: swept-area results are checked against an independent geometric metric and the cited planning machinery is external.

full rationale

I walked the derivation chain: A* seed -> MINCO smoothing (Eq. 1) -> obstacle/sweep optimization (Eq. 5) -> MPC tracking (Eq. 24) -> wheel-angle conversion (Eqs. 28-30) -> swept-area evaluation (Eq. 21). The only point that might look circular is Jsv: Eq. (16) minimizes heading-versus-velocity misalignment rather than the swept area S, but S is not defined through Jsv; it is computed separately from the swept-volume SDF of Eq. (21) with the Armijo search, so the reported Sexcess is an independent geometric measurement, not a refit of the objective. The MINCO/JE/JT gradient content is cited to [5] and [77], which are external works with no author overlap with this paper's author list. Self-citations in the bibliography are background (localization, odometry, multi-robot planning) and none are load-bearing for the central claim. Section IV-A states "due to hardware limitations, we rely on simulations to verify performance," a real external-validity limitation, and Eqs. (22) and (28) contain a sign/frame consistency risk, but these are correctness and validation concerns, not circular reductions. I found no step where a predicted quantity is identical to an input or fitted parameter by construction.

Assumptions & free parameters 9 free parameters · 6 assumptions · 0 invented entities

The framework rests on several hand-chosen weights and constraints, on a heuristic mapping from heading alignment to swept-area reduction, and on a simplified linear kinematic model. No concrete parameter values are reported, so the central quantitative result cannot be independently reconstructed from the text.

free parameters (9)
  • WE (energy cost weight)
    Used in Eqs. (1) and (5) to balance energy consumption in both optimization stages; value not reported.
  • WT (time cost weight)
    Used in Eqs. (1) and (5) to balance total time; value not reported.
  • WP (path deviation weight)
    Used in Eq. (1) to keep the smoothed trajectory close to the A* path; value not reported.
  • Wob (obstacle cost weight)
    Used in Eq. (5) to balance obstacle safety distance; value not reported.
  • Wsv (swept area cost weight)
    Used in Eq. (5) to balance the swept-area proxy; value not reported and crucial to the central claim.
  • d_th (safety distance threshold)
    Used in Eq. (12) to activate obstacle cost; value not reported.
  • MPC horizons Np and Nc
    Prediction and control horizons in Eq. (23); values not reported.
  • MPC weight matrices QQ and RR
    Weights in Eq. (24) balancing tracking error and control effort; values not reported.
  • Control and rate constraint bounds
    Umin, Umax, ΔUmin, ΔUmax in Eq. (27); values not reported.
assumptions (6)
  • domain assumption The multi-axle vehicle is treated as a single rigid rectangle with constant shape and no lateral tilt.
    Section III-A defines the vehicle area ξ and states lateral tilt is neglected; this is standard for top-view swept-area analysis but ignores chassis articulation.
  • domain assumption For any point, there exists a unique time t* at which the minimum distance to the swept area occurs, enabling Armijo line search.
    Section III-B3 invokes this uniqueness to compute the swept-area SDF; it is a smoothness assumption on the trajectory.
  • ad hoc to paper Minimizing the heading-velocity alignment penalty sum(Δφ)^2 minimizes the swept area S.
    Eqs. (15)-(16) define Jsv as a proxy for swept area, but the paper provides no derivation showing the proxy bounds or equals the true swept area.
  • ad hoc to paper The linear state-space model X(k+1)=X(k)+T u with body-frame velocities added directly to global coordinates is valid for MPC prediction.
    Eq. (22) omits the rotation matrix needed to convert body-frame Vx,Vy to global increments; this is a strong modeling assumption that appears incorrect for turning.
  • domain assumption Obstacles are static and represented by a precomputed signed distance field during planning.
    The SDF-based obstacle cost in Eqs. (9)-(12) assumes a fixed environment, which is not stated as a limitation.
  • standard math The gradients for energy and time costs are correctly adopted from prior work [5] without re-derivation.
    Section III-B1 states the gradients were rigorously derived in [5]; the paper trusts that background result.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Swept Volume-Aware Trajectory Planning and MPC Tracking for Multi-Axle Swerve-Drive AMRs." pith.science (2026). https://pith.science/paper/6SJ2ZTQJ

@misc{pith2026241216875,
  author       = {Pith},
  title        = {Pith review of: Swept Volume-Aware Trajectory Planning and MPC Tracking for Multi-Axle Swerve-Drive AMRs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6SJ2ZTQJ}},
  note         = {Machine review of arXiv:2412.16875}
}
read the original abstract

Multi-axle autonomous mobile robots (AMRs) are set to revolutionize the future of robotics in logistics. As the backbone of next-generation solutions, these robots face a critical challenge: managing and minimizing the swept volume during turns while maintaining precise control. Traditional systems designed for standard vehicles often struggle with the complex dynamics of multi-axle configurations, leading to inefficiency and increased safety risk in confined spaces. Our innovative framework overcomes these limitations by combining swept volume minimization with Signed Distance Field (SDF) path planning and model predictive control (MPC) for independent wheel steering. This approach not only plans paths with an awareness of the swept volume but actively minimizes it in real-time, allowing each axle to follow a precise trajectory while significantly reducing the space the vehicle occupies. By predicting future states and adjusting the turning radius of each wheel, our method enhances both maneuverability and safety, even in the most constrained environments. Unlike previous works, our solution goes beyond basic path calculation and tracking, offering real-time path optimization with minimal swept volume and efficient individual axle control. To our knowledge, this is the first comprehensive approach to tackle these challenges, delivering life-saving improvements in control, efficiency, and safety for multi-axle AMRs. Furthermore, we will open-source our work to foster collaboration and enable others to advance safer, more efficient autonomous systems.

Figures

Figures reproduced from arXiv: 2412.16875 by the authors.

Figure 1
Figure 1. This work aims to reduce the minimal swept volume and ensure stable trajectory tracking, enhancing safety in industrial applications.. The challenge lies in managing the swept volume while ensuring that each axle follows a safe and efficient path in real-time [4], [5]. Existing methods for path planning [8]– [19] and control [20]–[29] typically simplify the vehicle as a single rigid body [30], which does not account… view at source ↗
Figure 2
Figure 2. The proposed solution uses LiDAR inertial odometry [31] for front-end odometry, with multi-stage back-end planning and MPC to minimize swept volume iteratively. • We propose a method for calculating the steering angles of each wheel group in multi-axle vehicles based on velocity vectors, simplifying the vehicle model and facilitating the use of MPC control. • We validate the approach, showing significant reduc￾tions… view at source ↗
Figure 3
Figure 3. Vehicle Parametric Model and Tj , i.e., ∂Job/∂Tj and ∂Job/∂Pj , are calculated using the Signed Distance Field (SDF) [4]. In the top-down view as shown in [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Multi-Axle AMR MPC tracking control. distance occurs, the Armijo line search method [80] can be employed to find this optimal time t ∗ . By substituting t ∗ into Equation (21), we can obtain the minimum distance f ∗ SDF = FSDF(P, t∗ ) at a point P. To calculate the SDF…
Figure 5
Figure 5. Figure 5: Experimental results show the proposed MPC accurately tracks the planned trajectory while minimizing robot travel in LiDAR blind spots [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: Trajectory tracking error comparison. As shown in [PITH_FULL_IMAGE:figures/full_fig_p005_6.png]
Figure 7
Figure 7. Figure 7: Experiment Result, Proposed solution offers minimal swept volume compared to another baseline model. TABLE I: Comparison of Metrics for Different Methods Method Sexcess (m2 ) t (s) ey (m) eφ (°) Classic [1] 82.15 1.50 ±1.38 ±3.6 SVSDF [5] 75.62 2.5 ±0.04 ±0.04 Hierarch…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

80 extracted references · 68 canonical work pages

  1. [1]

    Modelling and dynamic tracking control of industrial vehicles with tractor-trailer structure,

    H. Zhao, Z. Liu, Z. Li, S. Zhou, W. Chen, C. Suo, and Y .-H. Liu, “Modelling and dynamic tracking control of industrial vehicles with tractor-trailer structure,” in 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) . IEEE, 2019

    work page 2019

  2. [2]

    Planning, learning and reasoning framework for robot truck unloading,

    F. Islam, A. Vemula, S.-K. Kim, A. Dornbush, O. Salzman, and M. Likhachev, “Planning, learning and reasoning framework for robot truck unloading,” in 2020 IEEE International Conference on Robotics and Automation (ICRA) . IEEE, 2020, pp. 5011–5017

    work page 2020

  3. [3]

    Survey on localization systems and algorithms for unmanned systems,

    S. Yuan, H. Wang, and L. Xie, “Survey on localization systems and algorithms for unmanned systems,” Unmanned Systems, vol. 9, no. 02, pp. 129–163, 2021

    2021

  4. [4]

    Continuous implicit sdf based any-shape robot trajectory optimization,

    T. Zhang, J. Wang, C. Xu, A. Gao, and F. Gao, “Continuous implicit sdf based any-shape robot trajectory optimization,” in 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) . IEEE, 2023, pp. 282–289

    2023

  5. [5]

    Implicit swept volume sdf: Enabling continuous collision- free trajectory generation for arbitrary shapes,

    J. Wang, T. Zhang, Q. Zhang, C. Zeng, J. Yu, C. Xu, L. Xu, and F. Gao, “Implicit swept volume sdf: Enabling continuous collision- free trajectory generation for arbitrary shapes,” ACM Transactions on Graphics (TOG), vol. 43, no. 4, pp. 1–14, 2024

    work page 2024

  6. [6]

    Experimental eval- uation of minimum swept-path control for autonomous reversing of articulated vehicles,

    X. Liu, A. K. Madhusudhanan, and D. Cebon, “Experimental eval- uation of minimum swept-path control for autonomous reversing of articulated vehicles,” in 2020 IEEE Intelligent Vehicles Symposium (IV). IEEE, 2020, pp. 584–589

    work page 2020

  7. [7]

    E2e parking: Autonomous parking by the end-to-end neural network on the carla simulator,

    Y . Yang, D. Chen, T. Qin, X. Mu, C. Xu, and M. Yang, “E2e parking: Autonomous parking by the end-to-end neural network on the carla simulator,” in 2024 IEEE Intelligent Vehicles Symposium (IV) . IEEE, 2024, pp. 2375–2382

    work page 2024

  8. [8]

    Multi-robot task planning under individual and collaborative temporal logic specifications,

    R. Bai, R. Zheng, M. Liu, and S. Zhang, “Multi-robot task planning under individual and collaborative temporal logic specifications,” in 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, 2021, pp. 6382–6389

    work page 2021

Show all 80 references
  1. [9]

    Risk-aware path planning using cvar for quadrotors,

    J. Bian, J. Zhang, K. Guo, W. Li, X. Yu, and L. Guo, “Risk-aware path planning using cvar for quadrotors,” in 2023 6th International Symposium on Autonomous Systems (ISAS) . IEEE, 2023, pp. 1–6

    2023

  2. [10]

    Gs-planner: A gaussian-splatting-based planning framework for active high-fidelity reconstruction,

    R. Jin, Y . Gao, Y . Wang, Y . Wu, H. Lu, C. Xu, and F. Gao, “Gs-planner: A gaussian-splatting-based planning framework for active high-fidelity reconstruction,” in 2024 IEEE/RSJ International Conference on Intel- ligent Robots and Systems (IROS) . IEEE, 2024, pp. 11 202–11 209

    2024

  3. [11]

    Safe maneuvering planning for flights in complex environments,

    S. Liu, K. Guo, X. Yu, L. Ma, L. Xie, and L. Guo, “Safe maneuvering planning for flights in complex environments,” IEEE Transactions on Industrial Electronics, vol. 71, no. 5, pp. 4944–4953, 2023

    2023

  4. [12]

    Path planning for multiple tethered robots using topological braids,

    M. Cao, K. Cao, S. Yuan, K. Liu, Y . L. Wong, and L. Xie, “Path planning for multiple tethered robots using topological braids,” in Robotics: Science and Systems , 2023

    2023

  5. [13]

    Neptune: non-entangling trajectory planning for multiple tethered unmanned vehicles,

    M. Cao, K. Cao, S. Yuan, T.-M. Nguyen, and L. Xie, “Neptune: non-entangling trajectory planning for multiple tethered unmanned vehicles,” IEEE Transactions on Robotics , vol. 39, no. 4, pp. 2786– 2804, 2023

    2023

  6. [14]

    Jacquard v2: Refining datasets using the human in the loop data correction method,

    Q. Li and S. Yuan, “Jacquard v2: Refining datasets using the human in the loop data correction method,” in 2024 IEEE International Conference on Robotics and Automation (ICRA), 2024, pp. 7932–7938

    2024

  7. [15]

    Design and anti-disturbance control for an h-configuration tiltable quadrotor: Enabling narrow space crossing,

    C. Zhao, E. Wang, K. Guo, and X. Yu, “Design and anti-disturbance control for an h-configuration tiltable quadrotor: Enabling narrow space crossing,” IEEE Transactions on Aerospace and Electronic Systems , 2024

    2024

  8. [16]

    Trace: Trajectory refinement with control error enables safe and accurate maneuvers,

    Z. Yang, J. Jia, Y . Liu, K. Guo, X. Yu, and L. Guo, “Trace: Trajectory refinement with control error enables safe and accurate maneuvers,” in 2024 IEEE 18th International Conference on Control & Automation (ICCA). IEEE, 2024, pp. 154–161

    2024

  9. [17]

    Flying in narrow spaces: Prioritizing safety with disturbance-aware control,

    D. Fan, Q. Liu, C. Zhao, K. Guo, Z. Yang, X. Yu, and L. Guo, “Flying in narrow spaces: Prioritizing safety with disturbance-aware control,” IEEE Robotics and Automation Letters , 2024

    2024

  10. [18]

    Aggressive collision-inclusive motion planning,

    H. Yu, C. Hu, J. Wang, G. Lu, J. Tu, Z. Zheng, J. Li, and F. Gao, “Aggressive collision-inclusive motion planning,” IEEE/ASME Trans- actions on Mechatronics , 2024

    2024

  11. [19]

    Realm: Real-time line-of-sight maintenance in multi-robot navigation with unknown obstacles,

    R. Bai, S. Yuan, K. Li, H. Guo, W.-Y . Yau, and L. Xie, “Realm: Real-time line-of-sight maintenance in multi-robot navigation with unknown obstacles,” in IEEE International Conference on Robotics and Automation (ICRA) , 2025

    2025

  12. [20]

    Bearing-ratio-of-distance rigidity theory with application to directly similar formation control,

    K. Cao, D. Li, and L. Xie, “Bearing-ratio-of-distance rigidity theory with application to directly similar formation control,” Automatica, vol. 109, p. 108540, 2019

    2019

  13. [21]

    Non-cooperative stochas- tic target encirclement by anti-synchronization control via range-only measurement,

    F. Liu, S. Yuan, W. Meng, R. Su, and L. Xie, “Non-cooperative stochas- tic target encirclement by anti-synchronization control via range-only measurement,” in 2023 IEEE International Conference on Robotics and Automation (ICRA) . IEEE, 2023, pp. 5480–5485

    2023

  14. [22]

    A stackelberg game approach for three- player autonomous racing,

    Z. Hu, X. Li, and M. Meng, “A stackelberg game approach for three- player autonomous racing,” in 2023 42nd Chinese Control Conference (CCC). IEEE, 2023, pp. 4886–4891

    2023

  15. [23]

    Online trajectory correction and tracking for facade inspection using autonomous uav,

    M. Cao, Y . Lyu, S. Yuan, and L. Xie, “Online trajectory correction and tracking for facade inspection using autonomous uav,” in 2020 IEEE 16th International Conference on Control & Automation (ICCA). IEEE, 2020, pp. 1149–1154

    2020

  16. [24]

    Multiple noncooper- ative targets encirclement by relative distance-based positioning and neural antisynchronization control,

    F. Liu, S. Yuan, W. Meng, R. Su, and L. Xie, “Multiple noncooper- ative targets encirclement by relative distance-based positioning and neural antisynchronization control,” IEEE Transactions on Industrial Electronics, vol. 71, no. 2, pp. 1675–1685, 2023

    2023

  17. [25]

    A cost-effective cooperative exploration and inspection strategy for heterogeneous aerial system,

    X. Xu, M. Cao, S. Yuan, T. H. Nguyen, T.-M. Nguyen, and L. Xie, “A cost-effective cooperative exploration and inspection strategy for heterogeneous aerial system,” in Proceedings of the 2024 IEEE In- ternational Conference on Control and Automation (ICCA) . IEEE, 2024

    2024

  18. [26]

    Development control and navigation of octocopter,

    M. J. Er, S. Yuan, and N. Wang, “Development control and navigation of octocopter,” in 2013 10th IEEE International Conference on Control and Automation (ICCA) . IEEE, 2013, pp. 1639–1643

    2013

  19. [27]

    Evolver: Online learning and prediction of disturbances for robot control,

    J. Jia, W. Zhang, K. Guo, J. Wang, X. Yu, Y . Shi, and L. Guo, “Evolver: Online learning and prediction of disturbances for robot control,” IEEE Transactions on Robotics , 2023

    2023

  20. [28]

    Distance-based multiple noncooperative ground target encirclement for complex envi- ronments,

    F. Liu, S. Yuan, K. Cao, W. Meng, and L. Xie, “Distance-based multiple noncooperative ground target encirclement for complex envi- ronments,” IEEE Transactions on Control Systems Technology, vol. 33, no. 1, pp. 261–273, 2025

    2025

  21. [29]

    Integration of sense and control for uncertain systems based on delayed feedback active inference,

    M. Ji, K. Pan, X. Zhang, Q. Pan, X. Dai, and Y . Lyu, “Integration of sense and control for uncertain systems based on delayed feedback active inference,” Entropy, vol. 26, no. 11, p. 990, 2024

    2024

  22. [30]

    Cyclops: Open platform for scale truck platooning,

    H. Lee, J. Park, C. Koo, J.-C. Kim, and Y . Eun, “Cyclops: Open platform for scale truck platooning,” in 2022 International Conference on Robotics and Automation (ICRA) . IEEE, 2022, pp. 8971–8977

    2022

  23. [31]

    Eigen is all you need: Efficient lidar-inertial continuous-time odome- try with internal association,

    T.-M. Nguyen, X. Xu, T. Jin, Y . Yang, J. Li, S. Yuan, and L. Xie, “Eigen is all you need: Efficient lidar-inertial continuous-time odome- try with internal association,” IEEE Robotics and Automation Letters , 2024

    2024

  24. [32]

    Deep prediction of swept volume geometries: Robots and resolutions,

    J. Baxter, M. R. Yousefi, S. Sugaya, M. Morales, and L. Tapia, “Deep prediction of swept volume geometries: Robots and resolutions,” in 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, 2020, pp. 6665–6672

    2020

  25. [33]

    Fast deep swept volume estimator,

    H.-T. L. Chiang, J. E. Baxter, S. Sugaya, M. R. Yousefi, A. Faust, and L. Tapia, “Fast deep swept volume estimator,” The International Journal of Robotics Research, vol. 40, no. 10-11, pp. 1068–1086, 2021

    2021

  26. [34]

    An optimisation-based path planner for truck-trailer systems with driving direction changes,

    P. Zips, M. B ¨ock, and A. Kugi, “An optimisation-based path planner for truck-trailer systems with driving direction changes,” in 2015 IEEE International Conference on Robotics and Automation (ICRA) . IEEE, 2015, pp. 630–636

    2015

  27. [35]

    Preview-based discrete-time dynamic formation control over directed networks via matrix-valued laplacian,

    K. Cao, X. Li, and L. Xie, “Preview-based discrete-time dynamic formation control over directed networks via matrix-valued laplacian,” IEEE Transactions on Cybernetics , vol. 50, no. 3, 2019

    2019

  28. [36]

    Vehicle swept path analysis based on gps data,

    V . Ili ´c, D. Gavran, S. Fric, F. Trp ˇcevski, and S. Vranjevac, “Vehicle swept path analysis based on gps data,” Canadian Journal of Civil Engineering, vol. 45, no. 10, pp. 827–839, 2018

    2018

  29. [37]

    Salient sparse visual odometry with pose-only supervision,

    S. Chen, K. Liu, C. Wang, S. Yuan, J. Yang, and L. Xie, “Salient sparse visual odometry with pose-only supervision,” IEEE Robotics and Automation Letters , vol. 9, no. 5, pp. 4774–4781, 2024

    2024

  30. [38]

    Deepdsair: Deep 6-dof camera relocalization using deblurred semantic-aware image representation for large-scale outdoor environments,

    M. A. Esfahani, K. Wu, S. Yuan, and H. Wang, “Deepdsair: Deep 6-dof camera relocalization using deblurred semantic-aware image representation for large-scale outdoor environments,” Image and Vision Computing, vol. 89, pp. 120–130, 2019

    2019

  31. [39]

    A fast and light-weight noniterative visual odometry with rgb-d cameras,

    Z. Yang, K. Xu, S. Yuan, and L. Xie, “A fast and light-weight noniterative visual odometry with rgb-d cameras,” Unmanned Systems, vol. 0, no. 0, pp. 1–13, 2024

    2024

  32. [40]

    M-divo: Multiple tof rgb-d cameras-enhanced depth–inertial–visual odometry,

    J. Xu, W. Yu, S. Huang, S. Yuan, L. Zhao, R. Li, and L. Xie, “M-divo: Multiple tof rgb-d cameras-enhanced depth–inertial–visual odometry,” IEEE Internet of Things Journal , vol. 11, no. 23, pp. 37 562–37 570, 2024

    2024

  33. [41]

    Robust rgb-d slam in dynamic environments for autonomous vehicles,

    T. Ji, S. Yuan, and L. Xie, “Robust rgb-d slam in dynamic environments for autonomous vehicles,” in 2022 17th International Conference on Control, Automation, Robotics and Vision (ICARCV) . IEEE, 2022, pp. 665–671

    2022

  34. [42]

    Graph optimality-aware stochastic lidar bundle adjustment with progressive spatial smoothing,

    J. Li, T.-M. Nguyen, M. Cao, S. Yuan, T.-Y . Hung, and L. Xie, “Graph optimality-aware stochastic lidar bundle adjustment with progressive spatial smoothing,” in arXiv preprint arXiv:2410.14565 , 2024

    2024 arXiv

  35. [43]

    A new approach to train convolutional neural networks for real-time 6-dof camera relocalization,

    M. A. Esfahani, K. Wu, S. Yuan, and H. Wang, “A new approach to train convolutional neural networks for real-time 6-dof camera relocalization,” in 2018 IEEE 14th international conference on control and automation (ICCA) . IEEE, 2018, pp. 81–85

    2018

  36. [44]

    I2ekf-lo: A dual-iteration extended kalman filter based lidar odometry,

    W. Yu, J. Xu, C. Zhao, L. Zhao, T.-M. Nguyen, S. Yuan, M. Bai, and L. Xie, “I2ekf-lo: A dual-iteration extended kalman filter based lidar odometry,” in 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , 2024, pp. 10 453–10 460

    2024

  37. [45]

    From local un- derstanding to global regression in monocular visual odometry,

    M. A. Esfahani, K. Wu, S. Yuan, and H. Wang, “From local un- derstanding to global regression in monocular visual odometry,” In- ternational Journal of Pattern Recognition and Artificial Intelligence , vol. 34, no. 01, p. 2055002, 2020

    2020

  38. [46]

    Pss-ba: Lidar bundle adjustment with progressive spatial smoothing,

    J. Li, T.-M. Nguyen, S. Yuan, and L. Xie, “Pss-ba: Lidar bundle adjustment with progressive spatial smoothing,” in 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , 2024, pp. 1124–1129

    2024

  39. [47]

    Robust loop closure by textual cues in challenging environments,

    T. Jin, X. Xu, Y . Yang, S. Yuan, T.-M. Nguyen, J. Li, and L. Xie, “Robust loop closure by textual cues in challenging environments,” IEEE Robotics and Automation Letters , vol. 10, no. 1, pp. 812–819, 2025

    2025

  40. [48]

    Multi- robot active graph exploration with reduced pose-slam uncertainty via submodular optimization,

    R. Bai, S. Yuan, H. Guo, P. Yin, W.-Y . Yau, and L. Xie, “Multi- robot active graph exploration with reduced pose-slam uncertainty via submodular optimization,” in 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , 2024, pp. 10 229–10 236

    2024

  41. [49]

    Hcto: Optimality-aware lidar inertial odometry with hybrid continuous time optimization for compact wearable mapping system,

    J. Li, S. Yuan, M. Cao, T.-M. Nguyen, K. Cao, and L. Xie, “Hcto: Optimality-aware lidar inertial odometry with hybrid continuous time optimization for compact wearable mapping system,” ISPRS Journal of Photogrammetry and Remote Sensing , vol. 211, pp. 228–243, 2024

    2024

  42. [50]

    Uloc: Learning to localize in complex large-scale environments with ultra- wideband ranges,

    T.-M. Nguyen, Y . Yang, T.-D. Nguyen, S. Yuan, and L. Xie, “Uloc: Learning to localize in complex large-scale environments with ultra- wideband ranges,” in IEEE International Conference on Robotics and Automation (ICRA), 2025

    2025

  43. [51]

    Ua-mpc: Uncertainty-aware model predictive control for motorized lidar odom- etry,

    J. Li, X. Xu, J. Liu, K. Cao, S. Yuan, and L. Xie, “Ua-mpc: Uncertainty-aware model predictive control for motorized lidar odom- etry,” in arXiv preprint arXiv:2412.13873 , 2024

    2024 arXiv

  44. [52]

    Outram: One-shot global localization via triangulated scene graph and global outlier pruning,

    P. Yin, H. Cao, T.-M. Nguyen, S. Yuan, S. Zhang, K. Liu, and L. Xie, “Outram: One-shot global localization via triangulated scene graph and global outlier pruning,” in 2024 IEEE International Conference on Robotics and Automation (ICRA) . IEEE, 2024, pp. 13 717–13 723

    2024

  45. [53]

    Sgba: Semantic gaussian mixture model-based lidar bundle adjustment,

    X. Ji, S. Yuan, J. Li, P. Yin, H. Cao, and L. Xie, “Sgba: Semantic gaussian mixture model-based lidar bundle adjustment,”IEEE Robotics and Automation Letters , vol. 9, no. 12, pp. 10 922–10 929, 2024

    2024

  46. [54]

    Learning to extract robust handcrafted features with a single observation via evolutionary neurogenesis,

    M. A. Esfahani, H. Wang, B. Bashari, K. Wu, and S. Yuan, “Learning to extract robust handcrafted features with a single observation via evolutionary neurogenesis,” Applied Soft Computing , vol. 106, p. 107424, 2021

    2021

  47. [55]

    Towards utilizing deep uncertainty in traditional slam,

    M. A. Esfahani, K. Wu, S. Yuan, and H. Wang, “Towards utilizing deep uncertainty in traditional slam,” in 2019 IEEE 15th International Conference on Control and Automation (ICCA) , 2019, pp. 344–349

    2019

  48. [56]

    Spins: A structure priors aided inertial navigation system,

    Y . Lyu, T.-M. Nguyen, L. Liu, M. Cao, S. Yuan, T. H. Nguyen, and L. Xie, “Spins: A structure priors aided inertial navigation system,” Journal of Field Robotics , vol. 40, no. 4, pp. 879–900, 2023

    2023

  49. [57]

    Real-time swept volume and distance computation for self collision detection,

    H. T ¨aubig, B. B ¨auml, and U. Frese, “Real-time swept volume and distance computation for self collision detection,” in 2011 IEEE/RSJ International Conference on Intelligent Robots and Systems , 2011, pp. 1585–1592

    2011

  50. [58]

    Autonomous object level segmentation,

    S. Yuan and H. Wang, “Autonomous object level segmentation,” in Proceedings of International Conference on Control, Automation, Robotics and Vision (ICARCV 2014) , 2014, pp. 33–37

    2014

  51. [59]

    An automatic self-calibration approach for wide baseline stereo cameras using sea surface images,

    H. Wang, W. Mou, X. Mou, S. Yuan, S. Ulun, S. Yang, and B.-S. Shin, “An automatic self-calibration approach for wide baseline stereo cameras using sea surface images,” Unmanned Systems, vol. 3, no. 04, pp. 277–290, 2015

    2015

  52. [60]

    Heterogeneous stereo: A human vision inspired method for general robotics sensing,

    H. Wang, S. Yuan, and K. Wu, “Heterogeneous stereo: A human vision inspired method for general robotics sensing,” in TENCON 2017-2017 IEEE Region 10 Conference . IEEE, 2017, pp. 793–798

    2017

  53. [61]

    Depth-based obstacle avoidance through deep reinforcement learning,

    K. Wu, M. A. Esfahani, S. Yuan, and H. Wang, “Depth-based obstacle avoidance through deep reinforcement learning,” in Proceedings of the 5th International Conference on Mechatronics and Robotics Engineer- ing, 2019, pp. 102–106

    2019

  54. [62]

    Safe autonomous exploration and adaptive path planning strategy using signed distance field,

    H. Wang, Y . Lin, W. Zhang, W. Ye, M. Zhang, and X. Dong, “Safe autonomous exploration and adaptive path planning strategy using signed distance field,” IEEE Access, vol. 11, 2023

    2023

  55. [63]

    Vision based autonomous uav plane estimation and following for building inspection,

    Y . Lyu, M. Cao, S. Yuan, and L. Xie, “Vision based autonomous uav plane estimation and following for building inspection,” in arXiv preprint arXiv:2102.01423, 2021

    2021 arXiv

  56. [64]

    Air-embodied: An efficient active 3dgs-based interaction and recon- struction framework with embodied large language model,

    Z. Qi, S. Yuan, F. Liu, H. Cao, T. Deng, J. Yang, and L. Xie, “Air-embodied: An efficient active 3dgs-based interaction and recon- struction framework with embodied large language model,” in arXiv preprint arXiv:2409.16019, 2024

    2024 arXiv

  57. [65]

    Hierarchical control strategies for multi-mode steering system of emergency rescue vehicle,

    F.-x. Xu, C. Zhou, and X.-h. Liu, “Hierarchical control strategies for multi-mode steering system of emergency rescue vehicle,” Mechatron- ics, vol. 85, p. 102834, 2022

    2022

  58. [66]

    Multi-objective collaborative control method for multi-axle distributed vehicle assisted driving,

    W. Wang, J. Li, X. Li, Z. Li, and N. Guo, “Multi-objective collaborative control method for multi-axle distributed vehicle assisted driving,” Applied Sciences, vol. 13, no. 13, p. 7769, 2023

    2023

  59. [67]

    Turning characteristic study of multi-axle compound steering vehicle,

    F. Gao and X.-y. Li, “Turning characteristic study of multi-axle compound steering vehicle,” in 2014 IEEE Conference and Expo Transportation Electrification Asia-Pacific (ITEC Asia-Pacific). IEEE, 2014, pp. 1–5

    2014

  60. [68]

    Dual-steering mode based on direct yaw moment control for multi-wheel hub motor driven vehicles: Theoretical design and experimental assessment,

    Z. Zhang, X.-j. Ma, C.-g. Liu, and S.-g. Wei, “Dual-steering mode based on direct yaw moment control for multi-wheel hub motor driven vehicles: Theoretical design and experimental assessment,” Defence Technology, vol. 18, no. 1, pp. 49–61, 2022

    2022

  61. [69]

    Control strategy of the multi- axle distributed mechanic and electric drive vehicle’s turning radius,

    J.-b. HU, M.-m. FU, X.-y. LI, and J. NI, “Control strategy of the multi- axle distributed mechanic and electric drive vehicle’s turning radius,” Transactions of Beijing institute of Technology , vol. 36, no. 11, pp. 1131–1135, 2016

    2016

  62. [70]

    On optimizing steering performance of multi-axle vehicle based on driving force control,

    Z. Wu and C. Chen, “On optimizing steering performance of multi-axle vehicle based on driving force control,” inMATEC Web of Conferences, vol. 124. EDP Sciences, 2017, p. 07005

    2017

  63. [71]

    Kinematic analysis of multi- axle steering system for articulated vehicle,

    S. Chaudhuri, V . Saini, and M. Singh, “Kinematic analysis of multi- axle steering system for articulated vehicle,” SAE Technical Paper, Tech. Rep., 2009

    2009

  64. [72]

    Learn to navigate au- tonomously through deep reinforcement learning,

    K. Wu, H. Wang, M. A. Esfahani, and S. Yuan, “Learn to navigate au- tonomously through deep reinforcement learning,” IEEE Transactions on Industrial Electronics , vol. 69, no. 5, pp. 5342–5352, 2021

    2021

  65. [73]

    Steering control strategies for a four- wheel-independent-steering bin managing robot,

    Y . Ye, L. He, and Q. Zhang, “Steering control strategies for a four- wheel-independent-steering bin managing robot,” IFAC-PapersOnLine, vol. 49, no. 16, pp. 39–44, 2016

    2016

  66. [74]

    Study on control schemes of flexible steering system of a multi-axle all-wheel-steering robot,

    P. Zhang, L. Gao, and Y . Zhu, “Study on control schemes of flexible steering system of a multi-axle all-wheel-steering robot,” Advances in Mechanical Engineering, vol. 8, no. 6, p. 1687814016651556, 2016

    2016

  67. [75]

    The steering performance analysis of multi-axle vehicle based on sideslip angle control strategy,

    P. X. Zhang, L. Gao, and Y . Q. Zhu, “The steering performance analysis of multi-axle vehicle based on sideslip angle control strategy,” Applied Mechanics and Materials , vol. 701, pp. 799–802, 2015

    2015

  68. [76]

    Steering mode research of five-axle all-wheel steering wheeled robots,

    P. Zhang, Y . Zhu, R. Huang, and X. Zhang, “Steering mode research of five-axle all-wheel steering wheeled robots,” Mechanical Design and Manufacturing, no. 3, pp. 39–42, 2015

    2015

  69. [77]

    Geometrically constrained tra- jectory optimization for multicopters,

    Z. Wang, X. Zhou, C. Xu, and F. Gao, “Geometrically constrained tra- jectory optimization for multicopters,” IEEE Transactions on Robotics, vol. 38, no. 5, pp. 3259–3278, 2022

    2022

  70. [78]

    lbfgs: Efficient l-bfgs and owl-qn optimization in r,

    A. Coppola and B. M. Stewart, “lbfgs: Efficient l-bfgs and owl-qn optimization in r,” 2020

    2020

  71. [79]

    Limited memory bundle method and its variations for large-scale nonsmooth optimization,

    N. Karmitsa, “Limited memory bundle method and its variations for large-scale nonsmooth optimization,” Numerical Nonsmooth Optimiza- tion: State of the Art Algorithms , pp. 167–199, 2020

    2020

  72. [80]

    Minimization of functions having lipschitz continuous first partial derivatives,

    L. Armijo, “Minimization of functions having lipschitz continuous first partial derivatives,” Pacific Journal of mathematics, vol. 16, no. 1, pp. 1–3, 1966

    1966

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.