Pith. sign in

REVIEW 4 major objections 6 minor 41 references

PREFAIL: Identifying Precursors to Failures in Robotic Lift-and-Place Tasks to Improve Task Execution Performance

T0 review · 4 major / 6 minor · reviewed 2026-08-01 · deepseek-v4-flash

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

Pith's one-line read Object drift predicts robot failures before the point of no return.

desk verdict The t_stop dataset and η metric are genuinely useful; the headline results are real but the baseline is degenerate and the real-world t_stop boundary needs reproducibility evidence. read the letter →

arxiv 2607.16921 v1 pith:HECEXUA4 submitted 2026-07-18 cs.RO cs.AI

Zeyu Shangguan , Rajas Chitale , Rutvik Patel , Satyandra K. Gupta , Daniel Seita This is my paper · ORCID
classification cs.ROcs.AI
keywords failurepredictionnon-prehensilemanipulationlift-and-placerelativemotionemergencystoptimingriskestimationroboticmaterialhandlingcounterfactuallabeling
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 argues that in high-speed non-prehensile lift-and-place, an impending failure shows up first as relative motion between the carried object and its carrier, and that this signal can be used to predict failures before the latest moment at which an emergency stop would still help. To make that claim testable, the authors introduce a labeling scheme that pinpoints this latest intervention time, t_stop, by replaying trajectories and applying emergency stops at successively earlier moments. They train a network that fuses multi-view images, relative-motion features, and robot state into a continuous risk value, and report high precision and recall in both simulation and real-world trials. They also introduce an intervention timeliness ratio, the fraction of the task completed when an alarm fires, and show their system reaches around 93–96% on that measure. The practical payoff is a robot that runs at top speed and slows only when danger approaches, rather than always creeping along.

What carries the argument

Three components carry the argument. First, historical backtracking: a binary-search procedure that replays a trajectory with emergency stops at candidate times to label t_stop, the last moment a stop prevents failure. Second, relative motion features: 2D ICP alignment of object and carrier contours in two camera views, producing a rotation and translation that quantifies the drift preceding a drop. Third, a two-branch network that fuses visual features, relative motion, and proprioceptive state (joint angles, joint speeds, end-effector position, and past risk values) into a risk score in [0,1]. The score is trained with both a classification loss (safe vs. past-threshold, calibrated so 0.5

What would settle it

Replay a set of real-world trajectories multiple times, applying the emergency stop at times around the labeled t_stop (e.g., t_stop-1, t_stop, t_stop+1) and record the fraction of drops. If the transition is not sharp — a stop at t_stop fails 30% of the time, or t_stop+1 succeeds 30% of the time — then the ground-truth labeling and the η metric are not measuring what they claim. A second test: train the network with an alternative boundary one time step earlier and check whether reported η changes systematically.

Watch

Extended reading notes

Core claim

The central claim is that the relative motion between the target object and the carrier — captured as rotation and translation differences via 2D iterative closest point on segmented contours — is the key precursor signal for failure in high-speed non-prehensile lift-and-place. The paper defines a precise ground truth: t_stop, the latest time at which an emergency stop prevents failure. PREFAIL is trained to output a continuous risk value where 0.5 marks t_stop. On simulation data it reaches precision 0.995, accuracy 0.997, recall 0.995, and intervention timeliness η=0.964; on real-world data the best average is precision 0.944, accuracy 0.963, recall 0.985, and η=0.932. The authors argue th

Load-bearing premise

The whole evaluation depends on t_stop being a clean, repeatable line: an emergency stop at that exact moment always prevents failure and one step later always fails, in both simulation and the real world, with no random variation in object placement or contact.

Editorial extensions

If this is right

  • If the relative-motion signal is as reliable as the experiments indicate, high-speed material handling can run near maximum speed and slow down only when risk rises, cutting cycle time without sacrificing safety.
  • The t_stop labeling protocol gives the field a concrete definition of an actionable prediction: a true positive is counted only if the alarm fires before t_stop, so future systems can be compared on timeliness as well as accuracy.
  • Because the predictor is fed from images and robot states rather than from a policy's action distribution, it should transfer across different control policies without retraining.
  • The same approach could extend to other non-prehensile operations such as pushing, tilting, or sweeping, as long as the moved object and carrier can be segmented.

Reading between the lines

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

  • The η metric rewards predicting exactly at t_stop; in deployment, any uncertainty would argue for firing a safety margin earlier, so the reported η≈0.93–0.96 should be read as an upper bound for a cautious system.
  • The method's reliance on clean contour segmentation invites a stress test: degrade segmentation quality (occlusion, lighting, motion blur) and measure how precision and η degrade; the paper lists this as a known limitation.
  • The t_stop concept could serve as a safety oracle for other time-critical robot actions, such as grasp aborts or collision-avoidance stops, wherever dynamics are deterministic enough for counterfactual replay.
  • A natural extension is sim-to-real transfer of the t_stop label: if a dynamics model can predict the boundary, the labor-intensive physical binary search could be replaced or augmented, enabling larger real-world datasets.
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 / 6 minor

Summary. The paper proposes PREFAIL, a policy-agnostic framework for predicting failure precursors in rapid lift-and-place manipulation. The method uses a two-branch network: one branch processes multi-view images and relative object-carrier motion from 2D ICP on segmented contours, while the other processes robot state and past risk values; the fused features are trained with a combination of classification and regression losses to output a continuous risk value in [0,1]. The main novel contribution is a data labeling pipeline, historical backtracking, that identifies t_stop, the latest time at which an emergency stop prevents failure, and t_fail, the failure time, with intermediate risk values assigned by Hermite spline interpolation. The paper evaluates PREFAIL on MuJoCo simulation (5,775 trajectories) and a real-world dataset (254 trajectories), reporting high precision, accuracy, recall, and an intervention timeliness ratio η. Results are compared against Fail-Detect, an OOD-based baseline with multiple variants, and extensive ablations are provided. The paper also includes a real-world deployment where risk-based velocity scaling prevents failure with small cycle-time overhead.

Significance. If the t_stop boundary is reproducible and the labeling scheme is sound, the paper makes a useful contribution: it is one of the first to attach a precise intervention deadline to each failure trajectory, enabling evaluation of not just whether a failure is detected but whether the detection is actionable. The relative-motion representation is simple, physically motivated, and policy-agnostic, and the real-world deployment demonstrates a concrete mechanism for trading cycle time against risk. The paper is unusually transparent in reporting full per-fold cross-validation tables and ablation studies, and the real-world dataset with t_stop labels is a potentially valuable resource. However, the significance is currently tempered by two concerns: the real-world t_stop boundary is treated as deterministic without repeatability evidence, and the only baseline is a degenerate OOD detector that, by the paper's own account, predicts every trajectory as failure. These issues do not invalidate the method, but they do mean that the headline claims of 'precision' and 'timeliness' are not yet fully supported.

major comments (4)
  1. [IV-A and Eq. (7)] The real-world t_stop boundary is load-bearing for the evaluation, since TP/FP-2 classification and η are all defined relative to t_stop. In Sec. IV-A, t_stop is found by physically replaying a rollout with e-stops at different times, but no repeatability statistics are reported. The outcome of an e-stop depends on precise resetting of the object and carrier state, and on the e-stop latency/timing jitter. If the binary search produces different t_stop on different replays, then Eq. (7) and the FP-2 'too late' classification are computed against a noisy reference. The paper should report, for a subset of real-world trajectories, repeated determinations of t_stop (e.g., 5–10 repetitions) with the resulting distribution and agreement at the t_stop vs t_stop+1 boundary. Without this, the claim that the dataset provides 'precise' intervention deadlines is not established.
  2. [IV-A, V-G, XII and Table III] The risk labels are constructed by anchoring 0 at t0, 0.5 at t_stop, and 1 at t_fail and interpolating with a Hermite spline. The network is then trained with a classification loss whose decision boundary is exactly the 0.5 level, and the evaluation uses the same 0.5 threshold to decide whether a prediction is before or after t_stop. This is not fully circular because t_stop is physically grounded, but the intermediate risk curve is an unverified supervisory signal. Table III shows that the choice of interpolation drastically changes the results (step interpolation collapses to precision 0.314, recall 1.0), so the reported η and precision/recall are partly properties of the chosen label scheme. The paper should provide a sensitivity analysis (e.g., varying the anchor values or using a label-free threshold selection) and, if possible, validate the interpolated risk curve against independe
  3. [V-B, Table I] The comparison to Fail-Detect is not yet convincing. As the paper itself notes, all Fail-Detect variants have precision equal to accuracy (0.37–0.51) and recall 1.0, which means they are equivalent to always predicting failure. The large margin in Table I is therefore against a degenerate baseline that provides no discrimination between failure and non-failure trajectories. The claim that PREFAIL 'substantially improves' accuracy and timeliness would be stronger if the paper included at least one non-degenerate baseline, such as a straightforward binary classifier trained on the same robot-state and image features, or a tuned version of Fail-Detect that trades recall for precision. Without such a baseline, it is unclear how much of the improvement comes from the proposed representations versus simply training a supervised classifier on well-balanced labels.
  4. [Table II] The ablation shows that removing past risk values causes the model to collapse to predicting every trajectory as failure (precision 0.314, accuracy 0.315, recall 1.0). This is concerning because past risk values are ground-truth during training and predicted during inference. If the model primarily propagates its previous risk estimates, then the high performance may be an artifact of autoregressive smoothing rather than genuine perception of failure precursors. The paper should report inference performance when the past-risk channel is fed only with the model's own predictions from a cold start, and also evaluate the model's first decision within a trajectory (where no past prediction exists). This would clarify whether the relative-motion and state features alone can support the claimed accuracy.
minor comments (6)
  1. [IV-A] The text describes the binary search as finding 'the earliest, safe intervention point,' but t_stop is the latest time at which an e-stop prevents failure. Please correct the wording to avoid confusion.
  2. [Eq. (2)] The symbol R is used both for rotation estimates R_o-c and for the risk value R_i. This is confusing in Eq. (2) and in the surrounding text. Use a distinct symbol (e.g., ρ) for the rotation component.
  3. [Eq. (7)] The definition η = t_pred_stop / t_stop can exceed 1.0 if the prediction occurs after t_stop, yet the text says the optimal η is 1.0 and that higher η is desirable. Please clarify whether η is defined only for true positives, capped at 1.0, or whether late predictions are excluded.
  4. [V-C] The description of the real-world evaluation is ambiguous: 'ten-fold cross-validation by constructing three different splits' could mean three different 10-fold splits, or something else. Please clarify the exact protocol and how the reported average in Table I is computed.
  5. [V-E] The caption of Fig. 5 says results are obtained under 'Sim&Real training and evaluation,' but Section V reports separate simulation and real-world experiments. Please clarify whether this refers to a mixed training condition or is a typo.
  6. [V-A] The paper says the real-world dataset contains 'two boxes with shifting contents,' which adds variability, but it does not report whether the t_stop annotation procedure was any different for these boxes. A brief note on how shifting contents affect the physical binary search would be helpful.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: t_stop is grounded by physical counterfactual replay, and PREFAIL is evaluated on held-out trajectories.

full rationale

The central labeling quantity t_stop is produced by an independent physical procedure—historical backtracking—in which trajectories are replayed with emergency stops at different times and the outcome is directly observed (Sec. IV-A). The model's risk predictions are trained against labels derived from t_stop and t_fail, but the evaluation is on held-out splits (simulation 8:1:1; real-world 10-fold cross-validation), so the reported precision/recall/η measure generalization rather than a fit to the training labels. The overlap between the classification loss (which uses labels 'after t_stop' vs 'before t_stop') and the η metric (t_pred_stop/t_stop) is standard supervised calibration, not a definitional reduction: t_stop is not defined by the model, and the model is not used to set t_stop. The paper explicitly acknowledges that intermediate risk values are interpolated rather than physically measured (Sec. V-G), which is a label-design choice and does not make the evaluation circular. The only self-citation ([40], used for box-shaped industrial objects) is not load-bearing for the main claims. The real-world t_stop reproducibility concern is a measurement-validity issue, not a circularity issue, and falls outside this pass's scope.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

No new physical entities are introduced; t_stop and η are operational labels/metrics with independent physical definitions for t_stop. The main burden is the set of domain assumptions about simulation fidelity, real-world repeatability, ICP accuracy, and the validity of synthetic risk labels, plus three explicit free parameters (m, α, label anchors/threshold).

free parameters (3)
  • temporal window length m = 5
    Set 'empirically' (Sec. IV, Problem Formulation); no sensitivity analysis is provided, and all branches use this window.
  • loss weight α = not reported
    Balances classification and regression losses in Eq. 4; the value is not given in the text, making the exact training objective unreproducible.
  • risk label anchors and threshold = 0.0, 0.5, 1.0
    Ground-truth risk is constructed as 0 at t0, 0.5 at t_stop, and 1 at t_fail, with Hermite interpolation in between; the same 0.5 value is used as the inference threshold (Sec. IV-A, Sec. V-G).
assumptions (5)
  • domain assumption MuJoCo contact and friction dynamics accurately model failure and e-stop outcomes for historical backtracking
    The simulation dataset and simulated t_stop labels rely on MuJoCo physics being faithful enough that binary search over e-stop times yields valid intervention boundaries (Sec. IV-A, Sec. XI).
  • domain assumption Real-world rollouts are repeatable enough for physical binary search of t_stop
    The real-world labeling replays the same trajectory many times and assumes e-stop outcomes are deterministic; no repeatability statistics are given (Sec. IV-A).
  • domain assumption 2D ICP over segmented contours accurately estimates relative object-carrier motion
    Relative motion features in Eq. 2-3 depend on 2D ICP accuracy; the paper validates this only visually in Appendix VIII and lists occlusion/lighting degradation as a limitation (Sec. V-G).
  • ad hoc to paper Hermite-spline interpolated risk values are a valid supervisory signal
    Intermediate risk values are not physically measured; the paper says they 'provide a relative ordering of risk' and chose Hermite over step/linear/Bezier based on downstream performance (Sec. V-G, Table III).
  • ad hoc to paper A fixed risk threshold of 0.5 consistently represents t_stop across object types and speeds
    The paper assumes the network learns a calibrated scale on which R>0.5 always means 'past the point of no return' without per-scenario tuning (Sec. V-G).

how reviews work

0 comments
Cite this review

Pith. "Pith review of PREFAIL: Identifying Precursors to Failures in Robotic Lift-and-Place Tasks to Improve Task Execution Performance." pith.science (2026). https://pith.science/paper/HECEXUA4

@misc{pith2026260716921,
  author       = {Pith},
  title        = {Pith review of: PREFAIL: Identifying Precursors to Failures in Robotic Lift-and-Place Tasks to Improve Task Execution Performance},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HECEXUA4}},
  note         = {Machine review of arXiv:2607.16921}
}
read the original abstract

Non-prehensile manipulation enables flexible material handling with part carriers, but friction-based support makes high-speed motions failure-prone, while slower operation increases cycle time. Proactive failure prediction is therefore essential for efficient and reliable performance, yet existing approaches remain limited by key constraints, including sensitivity to dynamic actions and high dependence on known policy structures. Furthermore, existing methods and datasets lack a precise characterization of the latest intervention time, leaving it unclear whether a detected failure can still be prevented through timely intervention. In this paper, we investigate lift-and-place tasks for non-prehensile material handling manipulation and propose a more effective approach to predicting precursors to failures (PREFAIL) by analyzing the relative motion of target objects with respect to the carrier. We further introduce a dataset that precisely identifies the latest intervention time for risky manipulations, enabling rigorous evaluation of whether a failure prediction is actionable. We validate our approach on both simulation and real-world datasets. Our experimental results demonstrate that PREFAIL substantially improves both the accuracy and timeliness of responses to failure precursors.

Figures

Figures reproduced from arXiv: 2607.16921 by the authors.

Figure 1
Figure 1. Our lift-and-place task involves a carrier (the pink plate in the [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Overview of PREFAIL (see Sec. IV for details). For each sample with m time steps, the observation branch processes multi-view images to extract visual features P1, then uses the segmented contours of the target object and the carrier to calculate their relative motion to get the motion features P2. The state branch uses E3 to encode robot states and past risk values to produce features P3. We fuse and process featur… view at source ↗
Figure 3
Figure 3. Historical backtracking data collection pipeline. The upper portion illustrates the iterative binary search process (across all iterations: Iter. 0, 1, 2, ..., n, n+1) for historically backtracking from the actual failure moment tfail (the black box drops from the pink panel), namely historical backtracking. In each iteration, we apply an e-stop (indicated by red border) and observe whether failure still occurs in t… view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Visualization of the objects and carriers used in our dataset. [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Visualization of all evaluation results on real-world test cases, with the predicted risk value curve (solid red line) vs the ground truth risk value d [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: Trajectory showing a real-world deployment scenario where joint [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: Reliability of 2D ICP. The contours from the last step are shown [PITH_FULL_IMAGE:figures/full_fig_p009_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

41 extracted references · 1 linked inside Pith

  1. [1]

    Mechanics and planning of manipulator pushing operations,

    M. T. Mason, “Mechanics and planning of manipulator pushing operations,”The International Journal of Robotics Research, 1986

    1986

  2. [2]

    Stable pushing: Mechanics, controllability, and planning,

    K. M. Lynch and M. T. Mason, “Stable pushing: Mechanics, controllability, and planning,”The international journal of robotics research, 1996

    1996

  3. [3]

    Nonprehensile dynamic manipula- tion: A survey,

    F. Ruggiero, V . Lippiello, and B. Siciliano, “Nonprehensile dynamic manipula- tion: A survey,”IEEE Robotics and Automation Letters (RA-L), 2018

    2018

  4. [4]

    Aha: A vision-language-model for detecting and reasoning over failures in robotic manipulation,

    J. Duan, W. Pumacay, N. Kumar, Y . R. Wang, S. Tian, W. Yuan, R. Krishna, D. Fox, A. Mandlekar, and Y . Guo, “Aha: A vision-language-model for detecting and reasoning over failures in robotic manipulation,” inInternational Conference on Learning Representations (ICLR), 2025

    2025

  5. [5]

    A multimodal han- dover failure detection dataset and baselines,

    S. Thoduka, N. Hochgeschwender, J. Gall, and P. G. Pl ¨oger, “A multimodal han- dover failure detection dataset and baselines,” inIEEE International Conference on Robotics and Automation (ICRA), 2024

    2024

  6. [6]

    Reactive or proactive? how robots should explain failures,

    G. LeMasurier, A. Gautam, Z. Han, J. W. Crandall, and H. A. Yanco, “Reactive or proactive? how robots should explain failures,” inInternational Conference on Human-Robot Interaction (HRI), 2024

    2024

  7. [7]

    Can We Detect Failures Without Failure Data? Uncertainty-Aware Runtime Failure Detection for Imitation Learning Policies,

    C. Xu, T. K. Nguyen, E. Dixon, C. Rodriguez, P. Miller, R. Lee, P. Shah, R. A. Ambrus, H. Nishimura, and M. Itkina, “Can We Detect Failures Without Failure Data? Uncertainty-Aware Runtime Failure Detection for Imitation Learning Policies,” inRobotics: Science and Systems (RSS), 2025

    2025

  8. [8]

    What matters in learning from offline human demonstrations for robot manipulation,

    A. Mandlekar, D. Xu, J. Wong, S. Nasiriany, C. Wang, R. Kulkarni, L. Fei- Fei, S. Savarese, Y . Zhu, and R. Mart´ın-Mart´ın, “What matters in learning from offline human demonstrations for robot manipulation,” inConference on Robot Learning (CoRL), 2022

    2022

Show all 41 references
  1. [9]

    Towards diverse behaviors: A benchmark for imitation learning with human demonstrations,

    X. Jia, D. Blessing, X. Jiang, M. Reuss, A. Donat, R. Lioutikov, and G. Neu- mann, “Towards diverse behaviors: A benchmark for imitation learning with human demonstrations,” inInternational Conference on Learning Representa- tions (ICLR), 2024

    2024

  2. [10]

    Diffusion policy: Visuomotor policy learning via action diffusion,

    C. Chi, Z. Xu, S. Feng, E. Cousineau, Y . Du, B. Burchfiel, R. Tedrake, and S. Song, “Diffusion policy: Visuomotor policy learning via action diffusion,” International Journal of Robotics Research (IJRR), 2025

    2025

  3. [11]

    Un- packing failure modes of generative policies: Runtime monitoring of consistency and progress,

    C. Agia, R. Sinha, J. Yang, Z. Cao, R. Antonova, M. Pavone, and J. Bohg, “Un- packing failure modes of generative policies: Runtime monitoring of consistency and progress,” inConference on Robot Learning (CoRL), 2025

    2025

  4. [12]

    Evaluating uncertainty-based failure detection for closed-loop llm planners,

    Z. Zheng, Q. Feng, H. Li, A. Knoll, and J. Feng, “Evaluating uncertainty-based failure detection for closed-loop llm planners,” inIEEE International Conference on Robotics and Automation (ICRA), 2024

    2024

  5. [13]

    Multimodal detection and classification of robot manipulation failures,

    A. Inceoglu, E. E. Aksoy, and S. Sariel, “Multimodal detection and classification of robot manipulation failures,”IEEE Robotics and Automation Letters (RA-L), 2024

    2024

  6. [14]

    Multimodal failure prediction for vision-based manipulation tasks with camera faults,

    Y . Ma, J. Liu, I. Mamaev, and A. Morozov, “Multimodal failure prediction for vision-based manipulation tasks with camera faults,” inIEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2024

    2024

  7. [15]

    Reflect: Summarizing robot experiences for failure explanation and correction,

    Z. Liu, A. Bahety, and S. Song, “Reflect: Summarizing robot experiences for failure explanation and correction,” inConference on Robot Learning (CoRL), 2023

    2023

  8. [16]

    Using large language models to generate and apply contingency handling procedures in collaborative assembly applications,

    J. H. Kang, N. Dhanaraj, S. Wadaskar, and S. K. Gupta, “Using large language models to generate and apply contingency handling procedures in collaborative assembly applications,” inIEEE International Conference on Robotics and Automation (ICRA), 2024

    2024

  9. [17]

    Multimodal coherent explanation generation of robot failures,

    P. Pramanick and S. Rossi, “Multimodal coherent explanation generation of robot failures,” inIEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2024

    2024

  10. [18]

    Racer: Rich language-guided failure recovery policies for imitation learning,

    Y . Dai, J. Lee, N. Fazeli, and J. Chai, “Racer: Rich language-guided failure recovery policies for imitation learning,” inIEEE International Conference on Robotics and Automation (ICRA), 2025

    2025

  11. [19]

    Real- Time Anomaly Detection and Reactive Planning with Large Language Models,

    R. Sinha, A. Elhafsi, C. Agia, M. Foutter, E. Schmerling, and M. Pavone, “Real- Time Anomaly Detection and Reactive Planning with Large Language Models,” inRobotics: Science and Systems (RSS), 2024

    2024

  12. [20]

    Future predictive success- or-failure classification for long-horizon robotic tasks,

    N. Sogi, H. Oyama, T. Shibata, and M. Terao, “Future predictive success- or-failure classification for long-horizon robotic tasks,” inInternational Joint Conference on Neural Networks (IJCNN), 2024

    2024

  13. [21]

    Code-as-monitor: Constraint-aware visual programming for reactive and proactive robotic failure detection,

    E. Zhou, Q. Su, C. Chi, Z. Zhang, Z. Wang, T. Huang, L. Sheng, and H. Wang, “Code-as-monitor: Constraint-aware visual programming for reactive and proactive robotic failure detection,” inIEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2025

    2025

  14. [22]

    Task failure prediction for wafer-handling robotic arms by using various machine learning algorithms,

    P. W. Huang and K.-J. Chung, “Task failure prediction for wafer-handling robotic arms by using various machine learning algorithms,”Measurement and Control, 2021

    2021

  15. [23]

    Failure prediction from limited hardware demonstrations,

    A. Parashar, K. Garg, J. Zhang, and C. Fan, “Failure prediction from limited hardware demonstrations,” in2025 61st Allerton Conference on Communication, Control, and Computing Proceedings, 2025

    2025

  16. [24]

    Grasp failure constraints for fast and reliable pick-and-place using multi-suction-cup grippers,

    J.-e. Lee, R. Sun, A. Bylard, and L. Sentis, “Grasp failure constraints for fast and reliable pick-and-place using multi-suction-cup grippers,”IEEE Transactions on Automation Science and Engineering, 2025

    2025

  17. [25]

    Uncovering robot vulnerabilities through semantic potential fields,

    S. Sagar, J. Duan, S. Vasudevan, Y . Zhou, H. B. Amor, D. Fox, and R. Senanayake, “Uncovering robot vulnerabilities through semantic potential fields,” inInternational Conference on Learning Representations (ICLR), 2026

    2026

  18. [26]

    Multi- task interactive robot fleet learning with visual world models,

    H. Liu, Y . Zhang, V . Betala, E. Zhang, J. Liu, C. Ding, and Y . Zhu, “Multi- task interactive robot fleet learning with visual world models,” inConference on Robot Learning (CoRL), 2025

    2025

  19. [27]

    Rediffuser: reliable decision-making using a diffuser with confidence estimation,

    N. He, S. Li, Z. Li, Y . Liu, and Y . He, “Rediffuser: reliable decision-making using a diffuser with confidence estimation,” inInternational Conference on Machine Learning (ICML), 2024

    2024

  20. [28]

    Consistency flow matching: Defining straight flows with velocity consistency,

    L. Yang, Z. Zhang, Z. Zhang, X. Liu, M. Xu, W. Zhang, C. Meng, S. Ermon, and B. Cui, “Consistency flow matching: Defining straight flows with velocity consistency,”arXiv preprint arXiv:2407.02398, 2024

    2024 arXiv

  21. [29]

    Natural posterior network: Deep bayesian predictive uncertainty for exponential family distributions,

    B. Charpentier, O. Borchert, D. Z ¨ugner, S. Geisler, and S. G ¨unnemann, “Natural posterior network: Deep bayesian predictive uncertainty for exponential family distributions,” inInternational Conference on Learning Representations (ICLR), 2022

    2022

  22. [30]

    Failure prediction at runtime for generative robot policies,

    R. R ¨omer, A. Kobras, L. Worbis, and A. P. Schoellig, “Failure prediction at runtime for generative robot policies,” inNeural Information Processing Systems (NeurIPS), 2025

    2025

  23. [31]

    Counterfactual reasoning: an analysis of in-context emergence,

    M. Miller, B. Sch ¨olkopf, and S. Guo, “Counterfactual reasoning: an analysis of in-context emergence,” inNeural Information Processing Systems (NeurIPS), 2025

    2025

  24. [32]

    Spline interpolation techniques,

    K. Erdogan, “Spline interpolation techniques,”Journal of Technical Science and Technologies, 2013

    2013

  25. [33]

    Method for registration of 3-d shapes,

    P. J. Besl and N. D. McKay, “Method for registration of 3-d shapes,” inSensor fusion IV: control paradigms and data structures, 1992

    1992

  26. [34]

    Go-icp: Solving 3d registration efficiently and globally optimally,

    J. Yang, H. Li, and Y . Jia, “Go-icp: Solving 3d registration efficiently and globally optimally,” inIEEE/CVF International Conference on Computer Vision (ICCV), 2013

    2013

  27. [35]

    Sam 2: Segment anything in images and videos,

    N. Ravi, V . Gabeur, Y .-T. Hu, R. Hu, C. Ryali, T. Ma, H. Khedr, R. R ¨adle, C. Rolland, L. Gustafson, E. Mintun, J. Pan, K. V . Alwala, N. Carion, C.-Y . Wu, R. Girshick, P. Dollar, and C. Feichtenhofer, “Sam 2: Segment anything in images and videos,” inInternational Confere...

    2025

  28. [36]

    Deep residual learning for image recognition,

    K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” inIEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016

    2016

  29. [37]

    Professor forcing: A new algorithm for training recurrent networks,

    A. M. Lamb, A. G. Alias Parth Goyal, Y . Zhang, S. Zhang, A. C. Courville, and Y . Bengio, “Professor forcing: A new algorithm for training recurrent networks,” Neural Information Processing Systems (NeurIPS), 2016

    2016

  30. [38]

    Flow matching for generative modeling,

    Y . Lipman, R. T. Chen, H. Ben-Hamu, M. Nickel, and M. Le, “Flow matching for generative modeling,” inInternational Conference on Learning Representations (ICLR), 2023

    2023

  31. [39]

    Mujoco: A physics engine for model- based control,

    E. Todorov, T. Erez, and Y . Tassa, “Mujoco: A physics engine for model- based control,” inIEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2012

    2012

  32. [40]

    Expanding Picking Actions for Time-Efficient Online 3D Bin Packing,

    N. Sarawgi, O. M. Manyar, F. Wang, T. H. Nguyen, D. Seita, and S. K. Gupta, “Expanding Picking Actions for Time-Efficient Online 3D Bin Packing,” inIEEE International Conference on Robotics and Automation (ICRA), 2026

    2026

  33. [41]

    Get a grip: Multi-finger grasp evaluation at scale enables robust sim-to-real transfer,

    T. G. W. Lum, A. H. Li, P. Culbertson, K. Srinivasan, A. Ames, M. Schwager, and J. Bohg, “Get a grip: Multi-finger grasp evaluation at scale enables robust sim-to-real transfer,” inConference on Robot Learning (CoRL), 2025. Appendix VII. DATAPOST-PROCESSING Our data post-proce...

    2025

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.