MPC-based Coarse-to-Fine Motion Planning for Robotic Object Transportation in Cluttered Environments

Chen Cai; Ernesto Dickel Saraiva; Steven Liu; Ya-Jun Pan

REVIEW 4 major objections 5 minor 23 references

MPC-based Coarse-to-Fine Motion Planning for Robotic Object Transportation in Cluttered Environments

T0 review · 4 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read A shrinking-horizon MPC with dual-camera perception and a refined kernel-perceptron collision detector lets two arms carry an object through cluttered, initially unmodeled scenes.

desk verdict A competent systems integration paper for vision-guided MPC with real hardware, but the robustness claims outrun the quantitative evidence. read the letter →

arxiv 2507.11211 v1 pith:BGLOF2LG submitted 2025-07-15 cs.RO cs.SYeess.SY

Chen Cai , Ernesto Dickel Saraiva , Ya-jun Pan , Steven Liu This is my paper · ORCID

classification cs.ROcs.SYeess.SY

keywords modelpredictivecontrolcoarse-to-fineplanningmulti-armmanipulationcollisionavoidancekernelperceptronB-splinetrajectoryoptimizationdual-cameraperceptionclosed-chainkinematics

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 presents a motion planner for a team of robot arms that must carry a rigid object through a cluttered room whose contents and exact target location are initially unknown. The planner starts from rough partial observations, produces a feasible global trajectory, and then continually refines the trajectory as new point-cloud data arrive from a stationary camera and a camera mounted on one arm's hand. The central claim is that this coarse-to-fine model predictive control (MPC) scheme, with a vision-cost term that steers the hand camera toward informative viewpoints and a kernel-based collision score enforced as a hard constraint, achieves whole-body collision avoidance and target-aware motion generation under uncertainty. Real-time experiments on a two-arm platform are used to support the claim, including a simulated human-intrusion case that triggers safety-aware replanning.

What carries the argument

The load-bearing mechanism is the pair formed by the B-spline trajectory transcription and the learnable collision score constraint. The full 21-dimensional state of both arms plus the object pose is encoded as a B-spline curve $z(s)$, so velocity and acceleration control points enter the optimization directly. The collision constraint is a refined kernel-perceptron proxy detector: the robot is split into separate collision groups, each with its own support vectors and a forward-kinematics kernel, and the optimization enforces $\mathrm{SCORE}_{g_i}(z(s_k)) = 0$ for each group at collocation points. A differentiable visibility cost $C_{\mathrm{vis}}$, built from reachable and unobstructed eye-in-hand camera poses, pulls the hand camera toward viewpoints that reveal the target, while the point-cloud pipeline (geometric filtering, convex-hull segmentation, occlusion polytopes) feeds the support-vector updates that keep the hard constraint current.

What would settle it

Place a solid, reachable obstacle between the arms and the target but hide it from both cameras at start (for example, just behind a larger box), then run the transport task. If the arms collide with the hidden obstacle before any perception update reveals it, the claim of whole-body safety in unmodeled environments collapses to safety against perceived obstacles only.

Watch

Extended reading notes

Core claim

The paper claims that cooperative transportation of a jointly held object in cluttered, unmodeled environments can be handled by a shrinking-horizon MPC in which the trajectory is encoded as B-splines and the environment is represented by a differentiable proxy collision score. At the core, the planner minimizes a cost over acceleration, joint-limit margin, duration, and a differentiable visibility score, subject to closed-chain kinematic constraints and the hard collision constraint $\mathrm{SCORE}(z(s_k)) = 0$ at every collocation point. The perception module fuses a stationary and an eye-in-hand depth camera, filters out the robot's own geometry, segments obstacles into convex hulls and occlusion polytopes, and continuously updates the support vectors of a kernel-perceptron collision detector so the hard constraint reflects newly seen obstacles. The authors report real-time applicability, whole-body collision avoidance, and dynamic replanning in experiments with two redundant 7-DoF cooperative robots, including exploratory eye-in-hand motion that resolves target visibility and a replanning mode triggered by a simulated human intrusion.

Load-bearing premise

Collision safety depends on perception completeness: every real obstacle must appear in the processed point cloud and be represented in the support-vector set, because the hard collision constraint can only reject configurations that the perceived geometry makes visible.

Editorial extensions

If this is right

The planner runs online at fixed replan intervals on a single workstation, so no pre-mapped environment is needed before the two arms start moving.
The differentiable visibility cost induces active exploration: the eye-in-hand camera is steered to resolve target pose, turning initial partial observations into precise placement.
The per-group collision scores plus active-learning updates let the hard constraints track slow scene changes, as shown by the human-intrusion scenario that triggers safety-aware evasive replanning.
Enforcing $h_{\mathrm{obj}}(q_{R1}) = h_{\mathrm{obj}}(q_{R2}) = x_{\mathrm{obj}}$ keeps the two arms coordinated during transport, and the near-identical independent pose estimates reported confirm closed-chain consistency.
Whole-body avoidance is enforced at collocation points, so not only the end effector but all link groups of both arms participate in the collision constraint.

Reading between the lines

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

The coarse-to-fine mechanism is not tied to two-arm transport: the same shrinking-horizon MPC with endpoint tolerance could let any manipulator reach into shelves or bins where the goal pose is only partially known, using the slack variable $\epsilon_{\mathrm{fp}}$ to represent goal uncertainty explicitly.
Because collision supports are learned independently per link group, a pre-trained boundary for the lower links might transfer across tasks that only change the wrist or gripper, potentially cutting retraining effort.
A testable extension is to add a penalty for trajectories that pass through still-unobserved regions rather than treating them as free space; that would directly sharpen the paper's weakest point, which is that the hard constraint only sees perceived obstacles.
The differentiable visibility score could serve outside planning as an objective for online view planning or for benchmarking active-perception systems, since it gives a smooth proxy for how reachable and unobstructed a camera pose is when looking at the target.

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.

Desk Editor's Note

read the letter

Dear colleague,

Quick take: this is a competent systems integration paper. What's new is the specific combination: dual-camera point cloud processing with incremental convex-hull and occlusion-polytope construction, a B-spline transcribed MPC that includes a learned kernel-perceptron collision constraint per link group, and a visibility cost that drives eye-in-hand camera poses. That combination is not in the cited literature, and the hardware results with two Franka arms show the pipeline works end-to-end.

The paper does several things well. The B-spline transcription with closed-chain constraints is clean. Splitting the DiffCo collision classifier into per-segment groups with separate support vectors is a sensible fix for the 'everything averaged together' problem, and the planar two-link comparison in Fig. 5 makes the point concretely. The visibility support vector idea is clever: it turns the problem of 'which camera pose sees the target' into a kernel similarity cost the MPC can optimize online.

The soft spots are real, though. The central claim is 'robust, whole-body collision avoidance under uncertainty,' but the experiments are single-run demonstrations without baselines, error bars, or repeated trials. The hard constraint SCORE(z(s_k))=0 is only as good as the learned support-vector set, which is trained from whatever the point-cloud pipeline currently sees. A missed or mis-segmented obstacle simply never enters the constraint. The paper does not impose a clearance margin, and checks collisions only at collocation points. That is a fundamental limitation of any perception-based planner, so the concern is not fatal, but it means the robustness claim should be scoped to 'obstacles that have been perceived and correctly segmented.' Right now the paper oversells it.

There is also circularity in the evaluation: the visibility score shown as evidence of exploration success is the same cost that the MPC optimizes. The task itself -- successfully placing the object -- is the real metric, and it would be easy to report. The absence of code or pseudocode until acceptance also keeps the failure modes uncheckable.

Who is this for? Robotics practitioners working on cooperative manipulation with partial observability. It deserves peer review as a systems letter, but it needs a revision that adds quantitative validation, a baseline comparison, and concrete statements about what the planner can and cannot guarantee. As written, I would not cite it as evidence of robust collision avoidance, only as an example of an integrated MPC-perception pipeline.

Recommendation: send to review, but require the experimental and code/reproducibility gaps to be addressed.

Referee Report

4 major / 5 minor

Summary. This manuscript proposes an MPC-based coarse-to-fine motion planning framework for two manipulators cooperatively transporting a rigid object through cluttered, initially unknown scenes. The planner uses a B-spline transcription of the combined joint/object state, enforces closed-chain and terminal-pose constraints, and incorporates a vision-based cost to drive an eye-in-hand camera toward informative views. The environment is represented by fusing stationary and eye-in-hand point clouds, applying geometric filtering, DBSCAN segmentation, and occlusion-polytope construction, while collision avoidance is encoded as hard constraints using a refined kernel-perceptron collision score with per-link-group support vectors and an active-learning update. The claims are evaluated with one static-clutter scenario and one human-intrusion scenario on a dual Franka Emika platform, with supplementary simulations in Drake.

Significance. If substantiated, the framework would be a useful integration of perception-driven MPC, learned collision constraints, and visibility-aware exploration for high-dimensional closed-chain manipulation, and the real dual-arm demonstration is commendable. Strengths include the explicit closed-chain formulation, the use of B-spline collocation with time scaling, the group-wise kernel and pruning refinements to DiffCo, and the deployment on real hardware with CasADi/IPOPT. However, the central claim of robust whole-body collision avoidance under uncertainty is not yet supported by the evidence: the evaluation is largely qualitative, no baselines or repeated trials are reported, and the safety-critical hard constraint rests on a learned surrogate whose completeness and accuracy are unquantified. The visibility evaluation is also partly circular because the reported score is the same kernel-similarity objective that the MPC maximizes.

major comments (4)

[IV-C, Eq. (6)-(10)] The hard constraint SCORE(z(s_k)) = 0 in Eq. (6) can only protect against obstacles that have been captured by the point-cloud pipeline of Section IV-B and converted into support vectors. Any obstacle that is missed, merged, or represented too coarsely never enters the support-vector set and hence cannot activate the constraint. In addition, the active-learning exploration step samples only a finite subset of the 14-DoF closed-chain configuration space, so even fully perceived obstacles may have false-negative regions in the learned score, and constraints are enforced only at collocation points with no clearance margin. As written, the paper does not certify 'robust, whole-body collision avoidance under uncertainty.' Please provide held-out collision-classification accuracy (false-negative rate versus clearance), an analysis of coverage of the closed-chain manifold, and a sensitivity study of the pruning and weight-reset parameters.
[V (Scenarios I and II)] The experimental section reports a single representative run per scenario with no repeated trials, no error bars, no comparison against baseline planners (e.g., the original DiffCo, a sampling-based planner, or an MPC without the visibility cost), and no quantitative success/failure statistics. The Abstract and Section I claim that the framework is 'experimentally validated' as robust, but the evidence is not commensurate with that claim. Please add multiple runs per scenario, report success rates, minimum distances to obstacles along the executed trajectory, planning and update times, and at least one baseline comparison.
[IV-D and Fig. 8] The top plot of Fig. 8 shows the visibility score computed as the kernel similarity to the 'visibility support vectors' of Section IV-D, which is exactly the quantity that the vision cost Cvis in Eq. (6) maximizes in the MPC. Reporting the increase of this score is therefore partly a check that the optimizer minimized its own objective, not independent evidence that exploration succeeded. Please evaluate visibility with an external metric, such as AprilTag detection success, target-pose estimation error from the raw point clouds, or pixel coverage of the target, and report target-localization accuracy at the end of exploration.
[IV-B, IV-C, IV-D] The manuscript repeatedly states that pseudocode 'will be released on Github upon acceptance,' leaving unspecified the active-learning reset rule, the Gram-matrix pruning thresholds, the biased-sampling distribution, the per-group support-vector limits, and the construction of 'visibility support vectors.' These are central algorithmic contributions, so as submitted the method is not fully reproducible and the identified failure modes cannot be independently checked. Please include complete pseudocode, parameter values, and ablation choices in the manuscript or an attached supplement.

minor comments (5)

[IV-A, Eq. (6)] The 'dexterity' cost term ||q(s_i) - (q_min + q_max)/2||^2 penalizes distance from the center of the joint range, but the prose says it steers agents clear of joint limits; the relationship between this cost and the hard limit constraints z_min <= c_i <= z_max should be clarified.
[V-B] The text says the top plot of Fig. 8 shows 'the visibility score defined in (6),' but Eq. (6) defines the optimal control problem, not the visibility score; please give the visibility score an explicit definition and equation number in Section IV-D.
[IV-C] The notation x is used for the Cartesian end-effector pose in Eq. (2) and for configuration vectors in the kernel definition in Section IV-C; the paper notes the conflict, but the repeated reuse of the symbol is still confusing to the reader.
[I and IV-A] There are several typos and grammatical issues, including 'Agent Dexterity: :' in the bullet list of Section IV-A, 'which not only expands the dimensionality...' fragment in the Introduction, and inconsistent comma use throughout.
[II] The sentence introducing GPMP2 cites reference [11], which is a survey; please cite the original GPMP2 paper so that readers can locate the method.

Circularity Check

1 steps flagged · score 4.0 of 10

One evaluation metric (visibility score) is the MPC objective itself, so the exploration 'validation' is partially circular; central collision and closed-chain claims remain independent.

self definitional [Section V-B (Fig. 8), Section IV-D, Eq. (6)]
"The top plot shows the visibility score defined in (6), computed as the similarity between the current eye-in-hand camera pose and the 'visibility support vectors' in IV-D. The plot is color-coded: red for partial visibility, green once sufficient visibility is achieved—demonstrating the exploratory behavior induced by our MPC planner."

The plotted 'visibility score' is the same similarity term that enters the MPC cost as WvisCvis in Eq. (6) via the 'visibility support vectors' of Section IV-D. The optimizer directly drives this quantity, so an increasing score is a restatement of the optimization objective rather than an independent measurement of target visibility. The paper does not report an external metric such as true positive target detection rate or pixel coverage; the snapshots provide only qualitative independent support. Thus the exploration-validation step reduces, by construction, to the cost being optimized.

full rationale

The central planning derivation (B-spline transcription, Eq. (6), closed-chain constraints, kernel-perceptron collision constraint) is not circular: the collision score is a learned surrogate trained on configuration-space labels, and the closed-chain and kinematic constraints are imposed independently. The paper's use of the authors' prior MPC formulation [16] and self-citations [13,23] is not load-bearing; the phase-variable representation is a standard coordinate choice. One genuine circularity appears in the experimental evaluation: the 'visibility score' used to demonstrate exploratory behavior in Fig. 8 is the same kernel-similarity cost that Eq. (6) optimizes, so that particular result is by construction. The real-world snapshots and closed-chain pose-disparity measurements give some independent content, and the collision-avoidance and replanning claims do not reduce to the fitted visibility proxy. The deferred pseudocode (Sections IV-B to IV-D) and unquantified perception-completeness assumption are significant correctness/safety risks but are not circularity.

Assumptions & free parameters 5 free parameters · 6 assumptions · 2 invented entities

The system relies on standard numerical optimization (B-spline transcription, IPOPT), on the closed-chain rigid-grasp model, and on the assumption that perceived point clouds capture all relevant obstacles. The method introduces several hand-chosen thresholds and weights whose values are not reported; there are no new physical entities, only computational constructs such as occlusion polytopes and visibility support vectors.

free parameters (5)

Objective weights W_a, W_m, W_d, W_vis, W_fp = not reported (hand-tuned)
Enter Eq. (6) and dominate the trade-off between smoothness, dexterity, duration, visibility, and terminal slack; values are not given.
B-spline degree n and control-point count M = not reported
Chosen in Eq. (4); these set trajectory resolution and smoothness, with no selection criterion given.
Support-vector limits and pruning thresholds per collision group = not reported
Section IV-C assigns higher limits to dynamic groups and adapts thresholds by collision status; no values are reported.
Active-learning sampling noise and biased sampling parameters = not reported
Section IV-C exploitation and exploration phases depend on sampling distributions and trajectory-based biasing; parameters are unspecified.
Kernel parameters for polyharmonic kernel = not reported
Used in Eq. (9) and Section IV-C; the kernel order affects the collision-score geometry and is not specified.

assumptions (6)

standard math B-spline basis functions have C^{n-1} continuity and satisfy the De Boor recursion used in Eqs. (4)-(5).
Standard spline theory, cited to [14,15].
domain assumption Convex hulls from DBSCAN and occlusion polytopes faithfully represent all obstacles that matter.
Section IV-B builds collision constraints from filtered point clouds; false negatives in segmentation remove obstacles from the SCORE constraint in Eq. (6).
domain assumption Rigid-grasp closed-chain equality h_obj(q_i) = x_obj holds throughout motion.
Section III-A; object dynamics are neglected because objects are lightweight. Violation would break the equality constraints in Eq. (6).
domain assumption The kernel-perceptron classifier SCORE generalizes from training support vectors to unseen configurations.
Section IV-C; the hard constraint SCORE = 0 is only as safe as classifier generalization, with no formal certificate provided.
ad hoc to paper Weight reset and support-vector pruning avoid numerical explosions without harming accuracy.
Section IV-C states raw active learning caused weight explosions; the fix is described without analysis or ablation.
ad hoc to paper Human intrusion can be represented by a tennis-ball proxy and by replanning mode with reset start and end constraints.
Section V-C; full human motion is not integrated, and [23] is cited as a future detector.

invented entities (2)

Occlusion polytope
purpose: Geometric object built from projected convex hull vertices to represent occluded regions for collision and visibility checks.
Internal computational construct from Section IV-B; no external validation beyond the reported experiments.
Visibility support vector
purpose: Set of camera poses with feasible inverse kinematics and unobstructed line of sight, used to define the differentiable vision cost C_vis in Eq. (6).
Constructed in Section IV-D by sampling and checking; no independent benchmark is provided.

how reviews work

0 comments

Cite this review

Pith. "Pith review of MPC-based Coarse-to-Fine Motion Planning for Robotic Object Transportation in Cluttered Environments." pith.science (2026). https://pith.science/paper/BGLOF2LG

@misc{pith2026250711211,
  author       = {Pith},
  title        = {Pith review of: MPC-based Coarse-to-Fine Motion Planning for Robotic Object Transportation in Cluttered Environments},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BGLOF2LG}},
  note         = {Machine review of arXiv:2507.11211}
}

read the original abstract

This letter presents a novel coarse-to-fine motion planning framework for robotic manipulation in cluttered, unmodeled environments. The system integrates a dual-camera perception setup with a B-spline-based model predictive control (MPC) scheme. Initially, the planner generates feasible global trajectories from partial and uncertain observations. As new visual data are incrementally fused, both the environment model and motion planning are progressively refined. A vision-based cost function promotes target-driven exploration, while a refined kernel-perceptron collision detector enables efficient constraint updates for real-time planning. The framework accommodates closed-chain kinematics and supports dynamic replanning. Experiments on a multi-arm platform validate its robustness and adaptability under uncertainties and clutter.

Figures

Figures reproduced from arXiv: 2507.11211 by the authors.

**Figure 2.** System overview with the MPC planner bridging task planning and [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** Left: FE cobot link indices and collision geometries [17]. Right: [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Figures from the paper (6 more)

**Figure 4.** Figure 4: Example configurations of a two-link planar robot. [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]

**Figure 5.** Figure 5: Comparison of original and modified algorithms in configuration [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]

**Figure 6.** Figure 6: Experimental setup. Main: Cooperative multi-arm platform with the [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]

**Figure 7.** Figure 7: Environment configuration and perceived workspace at trajectory [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]

**Figure 8.** Figure 8: Experimental results from Scenario I. (Top) Visibility score [PITH_FULL_IMAGE:figures/full_fig_p007_8.png]

**Figure 9.** Figure 9: Experimental results of Scenario II. Top: quantitative performance [PITH_FULL_IMAGE:figures/full_fig_p008_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

23 extracted references · 20 canonical work pages

[1]

Sampling-based motion planning: A comparative review,

A. Orthey, C. Chamzas, and L. E. Kavraki, “Sampling-based motion planning: A comparative review,” Annual Review of Control, Robotics, and Autonomous Systems , vol. 7, 2023

2023
[2]

Dynamic region-biased rapidly-exploring random trees,

J. Denny, R. Sandstr ¨om, A. Bregger, and N. M. Amato, “Dynamic region-biased rapidly-exploring random trees,” in Algorithmic F oun- dations of Robotics XII: Proceedings of the Twelfth Workshop on the Algorithmic F oundations of Robotics , pp. 640–655, Springer, 2020

work page 2020
[3]

Informed rrt*: Optimal sampling-based path planning focused via direct sampling of an admissible ellipsoidal heuristic,

J. D. Gammell et al. , “Informed rrt*: Optimal sampling-based path planning focused via direct sampling of an admissible ellipsoidal heuristic,” in Proc. IEEE/RSJ IROS , pp. 2997–3004, IEEE, 2014

work page 2014
[4]

Manipulation planning on constraint manifolds,

D. Berenson, S. S. Srinivasa, et al. , “Manipulation planning on constraint manifolds,” in Proc. IEEE ICRA , pp. 625–632, IEEE, 2009

work page 2009
[5]

Real-time obstacle avoidance for manipulators and mobile robots,

O. Khatib, “Real-time obstacle avoidance for manipulators and mobile robots,” Int. J. Robot. Res. , vol. 5, no. 1, pp. 90–98, 1986

1986
[6]

A dynamical system approach to realtime obstacle avoidance,

Khansari-Zadeh et al. , “A dynamical system approach to realtime obstacle avoidance,” Autonomous Robots, vol. 32, pp. 433–454, 2012

work page 2012
[7]

Neural potential field for obstacle-aware local motion planning,

M. Alhaddad et al. , “Neural potential field for obstacle-aware local motion planning,” in 2024 IEEE International Conference on Robotics and Automation (ICRA) , pp. 9313–9320, IEEE, 2024

work page 2024
[8]

Obstacle avoidance path planning of 6- dof robotic arm based on improved a* algorithm and artificial potential field method,

X. Tang, H. Zhou, and T. Xu, “Obstacle avoidance path planning of 6- dof robotic arm based on improved a* algorithm and artificial potential field method,” Robotica, vol. 42, no. 2, pp. 457–481, 2024

work page 2024

Show all 23 references

[9]

Reactive collision-free motion generation in joint space via dynamical systems and sampling- based mpc,

M. Koptev, N. Figueroa, and A. Billard, “Reactive collision-free motion generation in joint space via dynamical systems and sampling- based mpc,” Int. J. Robot. Res. , vol. 43, no. 13, pp. 2049–2069, 2024

2024
[10]

A dynamical system approach to decen- tralized collision-free autonomous coordination of a mobile assistive furniture swarm,

F. M. Conzelmann et al. , “A dynamical system approach to decen- tralized collision-free autonomous coordination of a mobile assistive furniture swarm,” in 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , pp. 7259–7265, IEEE, 2022

2022
[11]

Survey on recent advances in planning and control for collaborative robotics,

Y .-J. Pan, S. Buchanan, et al., “Survey on recent advances in planning and control for collaborative robotics,” IEEJ Journal of Industry Applications, vol. 14, no. 2, pp. 139–151, 2025

2025
[12]

Diffco: Autodifferentiable proxy collision detection with multiclass labels for safety-aware trajectory optimization,

Y . Zhi et al. , “Diffco: Autodifferentiable proxy collision detection with multiclass labels for safety-aware trajectory optimization,” IEEE Transactions on Robotics , vol. 38, no. 5, pp. 2668–2685, 2022

2022
[13]

Task space bilateral teleoper- ation of co-manipulators using power-based tdpc and leader-follower admittance control,

C. Cai, Y .-J. Pan, S. Liu, and L. Wan, “Task space bilateral teleoper- ation of co-manipulators using power-based tdpc and leader-follower admittance control,” in Proc. IEEE IECON , pp. 1–6, IEEE, 2021

2021
[14]

Continuous trajectory optimiza- tion via b-splines for multi-jointed robotic systems,

C. Wang, T. Xu, and M. Tomizuka, “Continuous trajectory optimiza- tion via b-splines for multi-jointed robotic systems,” arXiv preprint arXiv:2212.10500, 2022

2022 arXiv
[15]

B-spline based filters for multi- point trajectories planning,

L. Biagiotti and C. Melchiorri, “B-spline based filters for multi- point trajectories planning,” in 2010 IEEE International Conference on Robotics and Automation , pp. 3065–3070, IEEE, 2010

2010
[16]

A mpc-based approach for motion planning on redundant manipulators in human robot collaboration,

C. Cai, M. Schroeder, and S. Liu, “A mpc-based approach for motion planning on redundant manipulators in human robot collaboration,” in Proc. IEEE CASE , pp. 3387–3392, IEEE, 2024

2024
[17]

Drake: Model-based design and verification for robotics,

R. Tedrake and the Drake Development Team, “Drake: Model-based design and verification for robotics,” 2019

2019
[18]

A density-based algorithm for discovering clusters in large databases with noise,

M. Ester et al., “A density-based algorithm for discovering clusters in large databases with noise,” in kdd, vol. 96, pp. 226–231, 1996

1996
[19]

Learning-based proxy collision detection for robot motion planning applications,

N. Das and M. Yip, “Learning-based proxy collision detection for robot motion planning applications,” IEEE Transactions on Robotics , vol. 36, no. 4, pp. 1096–1114, 2020

2020
[20]

CasADi – A software framework for nonlinear optimization and optimal control,

J. A. E. Andersson, J. Gillis, G. Horn, et al. , “CasADi – A software framework for nonlinear optimization and optimal control,” Mathe- matical Programming Computation , vol. 11, no. 1, pp. 1–36, 2019

2019
[21]

Adaptive barrier update strategies for nonlinear interior methods,

J. Nocedal, A. W ¨achter, and R. A. Waltz, “Adaptive barrier update strategies for nonlinear interior methods,” SIAM Journal on Optimiza- tion, vol. 19, no. 4, pp. 1674–1693, 2009

2009
[22]

A distributed optimal control framework for multi-robot cooperative manipulation in dynamic environments,

Y . He, M. Wu, and S. Liu, “A distributed optimal control framework for multi-robot cooperative manipulation in dynamic environments,” Journal of Intelligent & Robotic Systems , vol. 105, no. 1, p. 8, 2022

2022
[23]

A probabilistic dynamic movement primitives framework on human hand motion prediction for an object transfer scenario,

C. Cai and S. Liu, “A probabilistic dynamic movement primitives framework on human hand motion prediction for an object transfer scenario,” IF AC-PapersOnLine, vol. 56, no. 2, pp. 8327–8332, 2023

2023

Pith tools

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

[1] [1]

Sampling-based motion planning: A comparative review,

A. Orthey, C. Chamzas, and L. E. Kavraki, “Sampling-based motion planning: A comparative review,” Annual Review of Control, Robotics, and Autonomous Systems , vol. 7, 2023

2023

[2] [2]

Dynamic region-biased rapidly-exploring random trees,

J. Denny, R. Sandstr ¨om, A. Bregger, and N. M. Amato, “Dynamic region-biased rapidly-exploring random trees,” in Algorithmic F oun- dations of Robotics XII: Proceedings of the Twelfth Workshop on the Algorithmic F oundations of Robotics , pp. 640–655, Springer, 2020

work page 2020

[3] [3]

Informed rrt*: Optimal sampling-based path planning focused via direct sampling of an admissible ellipsoidal heuristic,

J. D. Gammell et al. , “Informed rrt*: Optimal sampling-based path planning focused via direct sampling of an admissible ellipsoidal heuristic,” in Proc. IEEE/RSJ IROS , pp. 2997–3004, IEEE, 2014

work page 2014

[4] [4]

Manipulation planning on constraint manifolds,

D. Berenson, S. S. Srinivasa, et al. , “Manipulation planning on constraint manifolds,” in Proc. IEEE ICRA , pp. 625–632, IEEE, 2009

work page 2009

[5] [5]

Real-time obstacle avoidance for manipulators and mobile robots,

O. Khatib, “Real-time obstacle avoidance for manipulators and mobile robots,” Int. J. Robot. Res. , vol. 5, no. 1, pp. 90–98, 1986

1986

[6] [6]

A dynamical system approach to realtime obstacle avoidance,

Khansari-Zadeh et al. , “A dynamical system approach to realtime obstacle avoidance,” Autonomous Robots, vol. 32, pp. 433–454, 2012

work page 2012

[7] [7]

Neural potential field for obstacle-aware local motion planning,

M. Alhaddad et al. , “Neural potential field for obstacle-aware local motion planning,” in 2024 IEEE International Conference on Robotics and Automation (ICRA) , pp. 9313–9320, IEEE, 2024

work page 2024

[8] [8]

Obstacle avoidance path planning of 6- dof robotic arm based on improved a* algorithm and artificial potential field method,

X. Tang, H. Zhou, and T. Xu, “Obstacle avoidance path planning of 6- dof robotic arm based on improved a* algorithm and artificial potential field method,” Robotica, vol. 42, no. 2, pp. 457–481, 2024

work page 2024