Pith. sign in

REVIEW 4 major objections 5 minor 153 references

Efficiently Manipulating Clutter via Learning and Search-Based Reasoning

T0 review · 4 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read The dissertation argues that a learned push predictor combined with multi-step tree search solves clutter removal, object retrieval, and tabletop rearrangement with fewer actions and substantially less planning time.

desk verdict A competent dissertation that compiles five prior papers; useful as a readable summary of the author's line of work, but the claims about generality and speedups overstate what the evidence supports. read the letter →

arxiv 2505.08853 v1 pith:KDVZMJAH submitted 2025-05-13 cs.RO

Baichuan Huang This is my paper · ORCID
classification cs.RO
keywords robotmanipulationobjectretrievalfromclutterremovalpushpredictionMonteCarlotreesearchparallelsimulationself-supervisedlearningtabletoprearrangement
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

The dissertation is built around one proposal: before a robot touches a cluttered scene, it should imagine what each candidate push would do, then search several pushes ahead for the one sequence that makes the target graspable. The work develops a learned forward model, called the Deep Interaction Prediction Network (DIPN), that predicts how every object in the scene moves after a push and renders the predicted scene as an image a grasp-evaluation network can score. Wrapping this predictor in Monte Carlo tree search yields high completion rates in retrieval and rearrangement tasks while using fewer actions than reactive policies. The author reports that adding GPU-parallel batched simulation cuts planning time by more than 30 times relative to serial search, bringing decisions from minutes to a few seconds. If correct, the central claim is that learned interaction prediction plus long-horizon search, rather than either alone, is the practical recipe for real-time manipulation in clutter.

What carries the argument

The central mechanism is the learned forward model used as the transition function of a tree search. DIPN takes the current image, per-object masks, and a candidate push, and outputs a predicted translation and rotation for each object, then re-renders those movements into a synthetic post-push image. That image is scored by a grasp network, allowing the search to evaluate whether a push brings the target closer to graspable. Around this core, the thesis layers three search-side mechanisms: Visual Foresight Trees run MCTS with DIPN as the simulator; MORE uses the search's own Q-values as self-supervised labels to train a fast push-policy network (PPN) that guides later searches; and PMBS parallelizes MCTS by batching thousands of independent physics simulations on a GPU, using virtual loss to keep parallel selections from duplicating each other. The final chapter generalizes the same search architecture to actions that mix pick-and-place and push.

What would settle it

Run the same pipeline on a test set where objects have substantially different mass, friction, or deformability (for example cloth or crumpled paper) and where the target needs more than four pushes; if completion rates drop well below the 100% reported for wood blocks, the learned forward model and simulation-transfer assumptions are the weak link.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that object-interaction prediction and tree search are complementary: DIPN supplies a fast and legible model of push outcomes (over 90% average IoU on single-push prediction), and MCTS supplies the multi-step reasoning that turns those predictions into a minimal sequence of pushes before a grasp. The dissertation demonstrates the combination in escalating settings: clutter removal (DIPN+GN reaches 100% completion on hard instances in simulation), object retrieval (Visual Foresight Trees use 2.45 actions on average on 22 hard cases; MORE cuts planning time while matching solution quality; PMBS achieves over a 30x speedup with better solution quality than serial MCTS), and tabletop rearrangement with mixed push and pick-n-place primitives (PMMR completes 98% of simulated and 96.4% of real-robot cases). Real-robot tests are reported as transferring from simulation with minimal loss.

Load-bearing premise

The load-bearing premise is that the hand-built wood-block test cases and the physics simulator used for training and planning represent the real robot's world closely enough that actions chosen in simulation transfer to physical execution without modification.

Editorial extensions

If this is right

  • A robot can defer grasping until several pushes have created clearance, replacing greedy push-or-grasp decisions with planned sequences.
  • Because the forward model is learned from data rather than hand-tuned physics parameters, the same pipeline can transfer to new objects and to a physical robot with modest retraining.
  • Parallel batched simulation turns search from a minutes-per-decision bottleneck into a few-seconds-per-decision operation, making the approach viable in percept-plan-act loops.
  • Once a search policy is distilled into a fast push-policy network, planning cost falls further without sacrificing the number of actions used to complete a task.

Reading between the lines

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

  • Beyond the paper: if forward-model accuracy is the binding constraint, replacing the learned predictor with a differentiable simulator or training on more diverse object properties should extend the same search recipe to deformable and articulated objects.
  • Beyond the paper: the 30x speedup suggests that search-based manipulation planning is hardware-limited rather than algorithm-limited; as GPU simulation grows cheaper, the same algorithm should scale to re-planning in dynamic scenes.
  • Beyond the paper: MORE's search-then-distill loop could be iterated indefinitely, using a bootstrapped policy to make search deeper and generate better training labels; a direct test would measure solution quality as a function of the number of search-and-train rounds.
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 dissertation develops a sequence of algorithms for robotic manipulation in clutter, centered on integrating learned push prediction with search-based planning: DIPN for single-step push outcome prediction, VFT (an MCTS-DIPN hybrid) for object retrieval, MORE for self-supervised learning-guided MCTS, PMBS for GPU-parallel MCTS with batched rigid-body simulations, and HBFS/PMMR for multi-primitive tabletop rearrangement. The central claims are that DIPN achieves over 90% IoU accuracy in push prediction, that the integrated systems attain 100% completion on challenging retrieval scenarios with fewer actions than prior baselines, and that PMBS attains over 30x planning speedups while maintaining or improving solution quality. Evaluation is performed in PyBullet, CoppeliaSim, and Isaac Gym, and on a UR5e with a Robotiq 2F-85 gripper and a RealSense camera.

Significance. If the stated results hold, the thesis demonstrates a credible route to combining learned interaction models with look-ahead search for long-horizon manipulation planning, and the PMBS framework addresses a genuine computational bottleneck in MCTS-based planning. The work includes commendable elements: fully self-supervised data collection for the perception and prediction networks, detailed pseudo-code, extensive ablation studies, and real-robot experiments that go beyond the simulation benchmarks. However, the headline comparisons and the generalization claims need to be substantially tightened before the contributions can be accepted at face value. In particular, the 'over 30x speedup' claim rests on an apples-to-oranges comparison across different time budgets, and the baseline comparisons rely on numbers quoted from prior papers under different experimental setups.

major comments (4)
  1. [§4.4.1, Table 4.1, Figure 4.5] The test-case counts are internally inconsistent. Section 4.4.1 states that the test set includes 'the full set of 14 test cases from [48]' plus 18 hand-designed cases (32 total), yet Figure 4.5 and Table 4.1 report on '10 test cases from [48]', and the text later says the harder set contains '18 manually designed instances and 4 cases from [48]' (22 total). Because Tables 4.1 and 5.2 compare against go-PGN results quoted from [48], the reader cannot determine whether the comparison is on the same cases. Please reconcile these counts and specify exactly which cases were used for each comparison.
  2. [§6.4.1, Figure 6.8] The 'over 30x speedup' claim in the abstract and Chapter 6 is not supported by the experimental design as reported. The 855/28 = 30x figure compares PMBS with a 15-second budget (28s mean planning time) to serial MCTS with a 480-second budget (855s mean planning time). At matched time budgets, Table 6.1 reports 35s versus 301s for PMBS-60 versus MCTS-60, an 8.6x speedup. The 30x claim therefore conflates different operating points and should either be reported as a matched-solution-quality comparison at different budgets or be replaced with the matched-budget number.
  3. [§6.4.2, Appendix A] The claim of 'minimal sim-to-real performance loss' is based on only six hand-selected cases with known object models and pose estimation (Table 6.2), and the two appendix case studies are anecdotal, with one showing a case where 'the simulator does not provide accurate physics'. Since the dissertation's broader framing claims applicability to 'unstructured real-world settings' (Chapter 1 and Chapter 8), the generalization claim is load-bearing and is not established. Please either provide a quantitative evaluation across a more diverse set of objects, frictions, and arrangements, or explicitly narrow the scope of the conclusions to the tested regime.
  4. [§4.4.3, §5.4.1] The comparisons to gc-VPG and go-PGN rely on results quoted from [48] rather than local reimplementation, under experimental differences that include a 13cm versus 5cm effective push distance, a different gripper (RD2 versus 2F-85), and a different simulator (CoppeliaSim versus PyBullet). The statement that these differences 'do not provide our algorithm an unfair advantage' is not supported by any sensitivity analysis. Because the state-of-the-art claim depends on these numbers, please either reimplement the baselines locally or provide a quantitative discussion of how the setup differences could bias the comparison.
minor comments (5)
  1. [Table 6.1] The header says 'Time budgets are limited up to 60 seconds', yet the PMBS-60 (c=0) row reports a planning time of 113 seconds. Please clarify whether the budget applies per decision step or per episode, and explain why the mean planning time can exceed the stated budget.
  2. [Tables 3.1, 4.1-4.4, 5.1-5.3, 6.1-6.2] Most aggregate tables report only mean values without variance or confidence intervals. Given that the accompanying figures show nontrivial trial-to-trial variability, please add error bars, standard deviations, or per-trial data for the key metrics.
  3. [§4.4.1, §6.3.2] The grasp threshold is set to 0.8 in simulation and 0.7 in the real experiments in Chapter 4, and the grasp classifier threshold R*_c is set to 0.9 in Chapter 6. The sensitivity of the results to these hand-picked thresholds is not analyzed. A brief threshold sensitivity study would strengthen the reproducibility of the results.
  4. [§7.4.3, Table 7.3] The footnote states that robot time is 'recorded only for successful cases', which makes the comparison between PMMR-40 and HBFS on execution time hard to interpret. Please report the number of successful trials behind each mean and, ideally, the distribution of execution times.
  5. [§6.1] The introduction to Chapter 6 refers to 'MoJuCo' where the intended simulator name is 'MuJoCo'. Please correct this typo.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the dissertation's predictions and planning claims are validated against independent simulation and real-robot outcomes, not against their own inputs.

full rationale

The derivation chain is empirical rather than definitional. DIPN is trained on random-push data and evaluated by IoU against ground-truth post-push states (Section 3.4.1); GN is trained on grasp success/failure and used only as a reward estimator, with final task success measured by completion, grasp success, and action efficiency (Section 3.4.3). VFT uses the learned DIPN as a transition model inside MCTS, but terminal rewards come from GN on predicted states and are back-propagated; the benchmark cases are independent hand-designed arrangements (Figure 4.4). Chapter 5's PPN is a student network trained on MCTS Q-values and then used as a search prior; this is self-distillation, not circular reduction, because the reported metrics (number of actions, planning time, completion, grasp success) are evaluated by executing or simulating the resulting actions and are not equal to PPN's training labels by construction. PMBS speedups are wall-clock comparisons against a serial MCTS baseline using identical hardware, with solution quality measured by external simulation and robot execution. The only self-citations are to the author's own prior chapters, but those chapters are reproduced in the dissertation and the claims do not rest on an unverified uniqueness theorem or on a fitted parameter being renamed as a prediction. Acknowledged limitations (known object models, wood-block test corpus, small sim-to-real gap) are generality concerns, not circularity.

Assumptions & free parameters 7 free parameters · 5 assumptions · 0 invented entities

The central claims rest on multiple hand-chosen hyperparameters, strong domain assumptions (rigid objects, quasi-static pushes, known object models), and the assumption that simulation is faithful to the real world. No fundamentally new entities are introduced.

free parameters (7)
  • Discount factor gamma = 0.9 (Ch3), 0.8 (Ch4, Ch6), 0.5 (Ch5), 0.9 (Ch7)
    Chosen to penalize long action sequences; values differ across chapters without a sensitivity study.
  • Grasp threshold R*_g = 0.8 simulation, 0.7 real (Ch4); 0.7 (Ch3)
    Used to decide whether to grasp directly or push first; hand-tuned and different between simulation and real.
  • MCTS iteration budget Nmax = 150 (Ch4), 300 (Ch5 data collection), 50 (Ch5 eval), 10 (Ch5 real), 60s (Ch6), 40s (Ch7)
    Search budget affects solution quality and planning time; chosen per experiment without systematic optimization.
  • Push distance = 5 cm effective (Ch4), 10 cm (Ch5), 5 cm (Ch6), 7.4 cm training (Ch4)
    Action primitive length, set by hand.
  • UCB exploration constant C = 2 (Ch4, Ch5), 0.3 (Ch6), 1.5 (Ch7)
    Exploration-exploitation balance; no ablation or tuning report.
  • Top-m rollouts m = 3 (Ch4), 10 (Ch5), 100 (Ch7)
    MCTS Q-value uses the best m returns rather than the average.
  • Reward shaping parameters (ro, Rg, beta, gamma) = ro=0.7, Rg=2roN, beta=0.5, gamma=0.9 (Ch7)
    Heuristic reward weights for the rearrangement reward function; handpicked.
assumptions (5)
  • domain assumption Objects are rigid and manipulation primitives are quasi-static.
    Stated in Ch4 Section 4.2.1 ('The objects are rigid') and Ch7 Section 7.2.1 ('all primitives are considered to be quasi-static').
  • domain assumption Simulation (PyBullet, Isaac Gym) and learned models transfer to the real robot with negligible sim-to-real gap.
    Ch6 reports 'minimal sim-to-real differences'; this is an assumption for the real-world claims.
  • domain assumption Object models and poses are known for planning.
    Ch5 summary states 'we need to know the models of the objects to do the planning'; Ch7 uses pose estimation but assumes known object geometry.
  • ad hoc to paper Hand-designed test cases are representative of the target tasks.
    Success rates are measured on 10-25 manually created cases, which are reused across chapters; no standard benchmark is used.
  • standard math Standard supervised learning assumptions (i.i.d. training data, generalization to test scenes).
    Neural network training assumes the usual ML setup, without explicit verification of distribution shift.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Efficiently Manipulating Clutter via Learning and Search-Based Reasoning." pith.science (2026). https://pith.science/paper/KDVZMJAH

@misc{pith2026250508853,
  author       = {Pith},
  title        = {Pith review of: Efficiently Manipulating Clutter via Learning and Search-Based Reasoning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KDVZMJAH}},
  note         = {Machine review of arXiv:2505.08853}
}
read the original abstract

This thesis presents novel algorithms to advance robotic object rearrangement, a critical task for autonomous systems in applications like warehouse automation and household assistance. Addressing challenges of high-dimensional planning, complex object interactions, and computational demands, our work integrates deep learning for interaction prediction, tree search for action sequencing, and parallelized computation for efficiency. Key contributions include the Deep Interaction Prediction Network (DIPN) for accurate push motion forecasting (over 90% accuracy), its synergistic integration with Monte Carlo Tree Search (MCTS) for effective non-prehensile object retrieval (100% completion in specific challenging scenarios), and the Parallel MCTS with Batched Simulations (PMBS) framework, which achieves substantial planning speed-up while maintaining or improving solution quality. The research further explores combining diverse manipulation primitives, validated extensively through simulated and real-world experiments.

Figures

Figures reproduced from arXiv: 2505.08853 by the authors.

Figure 4
Figure 4. [PITH_FULL_IMAGE:figures/full_fig_p012_4.png] view at source ↗
Figure 5
Figure 5. [PITH_FULL_IMAGE:figures/full_fig_p012_5.png] view at source ↗
Figure 4
Figure 4. [PITH_FULL_IMAGE:figures/full_fig_p016_4.png] view at source ↗
Figures from the paper (90 more)
Figure 4
Figure 4. Figure 4 [PITH_FULL_IMAGE:figures/full_fig_p017_4.png]
Figure 7
Figure 7. Figure 7 [PITH_FULL_IMAGE:figures/full_fig_p018_7.png]
Figure 1
Figure 1. Figure 1 [PITH_FULL_IMAGE:figures/full_fig_p019_1.png]
Figure 1.1
Figure 1.1. Figure 1.1: Examples of robot manipulation tasks. (a) Grasping objects from clutter: This [PITH_FULL_IMAGE:figures/full_fig_p020_1_1.png]
Figure 1.2
Figure 1.2. Figure 1.2: Structure of the dissertation. Chapter 3 introduces the Deep Interaction Prediction Network for one-step push prediction in clutter removal. Chapter 4 extends to multi-step planning for efficient object retrieval using push actions. Chapters 5 and 6 explore GPU￾accel…
Figure 3
Figure 3. Figure 3: c [PITH_FULL_IMAGE:figures/full_fig_p032_3.png]
Figure 3.1
Figure 3.1. Figure 3.1: (a) The system setup includes a workspace with objects to remove, a Universal [PITH_FULL_IMAGE:figures/full_fig_p033_3_1.png]
Figure 3
Figure 3. Figure 3: a [PITH_FULL_IMAGE:figures/full_fig_p033_3.png]
Figure 3
Figure 3. Figure 3 [PITH_FULL_IMAGE:figures/full_fig_p035_3.png]
Figure 3.2
Figure 3.2. Figure 3.2: DIPN flow with an example. The network components dedicated to an object are color-coded to match the object. We only show the full network for the blue triangle object; the instance-specific structures for the other objects share the same weights and are simplified …
Figure 3.3
Figure 3.3. Figure 3.3: Sampled action in purple arrows around each object. [PITH_FULL_IMAGE:figures/full_fig_p036_3_3.png]
Figure 3
Figure 3. Figure 3 [PITH_FULL_IMAGE:figures/full_fig_p037_3.png]
Figure 3
Figure 3. Figure 3 [PITH_FULL_IMAGE:figures/full_fig_p039_3.png]
Figure 3.4
Figure 3.4. Figure 3.4: Architecture of GN. Pink, blue, and green text are used for channel count, image size, and kernel size, respectively. In training GN, image-based pre-training [103] was employed. The pre-training process treats pixel-wise grasping as a vision task to obtain a good ne…
Figure 3.5
Figure 3.5. Figure 3.5: DIPN learning curve with standard deviation shown as shaded regions. The x-axis is the number of pushes for training DIPN. The y-axis is the prediction error: 1 − IoU. The dotted and dashed lines are baselines. DIPN is compared with two baselines: the first one, call…
Figure 3
Figure 3. Figure 3 [PITH_FULL_IMAGE:figures/full_fig_p042_3.png]
Figure 3.6
Figure 3.6. Figure 3.6: shows typical predictions by DIPN. The network is learned in simulation with randomly shaped and colored objects, and directly transferred to the real system. We observe that DIPN can accurately predict the state after a push, with good accuracy on object orientation…
Figure 3.7
Figure 3.7. Figure 3.7: Manually generated hard instances largely similar to the ones in [ [PITH_FULL_IMAGE:figures/full_fig_p043_3_7.png]
Figure 3
Figure 3. Figure 3 [PITH_FULL_IMAGE:figures/full_fig_p044_3.png]
Figure 3.8
Figure 3.8. Figure 3.8: Grasp learning curves of algorithms for PaG in simulation. The x-axis is the total number of training steps, i.e., number of actions taken, including push and grasp. The y-axis is the grasp success rate. The dashed lines denote the success rate for a grasp right afte…
Figure 3
Figure 3. Figure 3: a [PITH_FULL_IMAGE:figures/full_fig_p045_3.png]
Figure 3.9
Figure 3.9. Figure 3.9: Grasp learning curves for PaG in real experiment. Solid lines indicate grasp success rate and dotted lines indicate push-then-grasp success rates over training steps. The GN is trained in a grasp only manner [PITH_FULL_IMAGE:figures/full_fig_p046_3_9.png]
Figure 4.1
Figure 4.1. Figure 4.1: (a) The hardware setup for object retrieval in a clutter includes a Universal [PITH_FULL_IMAGE:figures/full_fig_p049_4_1.png]
Figure 4
Figure 4. Figure 4 [PITH_FULL_IMAGE:figures/full_fig_p049_4.png]
Figure 4
Figure 4. Figure 4: a [PITH_FULL_IMAGE:figures/full_fig_p050_4.png]
Figure 4.2
Figure 4.2. Figure 4.2: Overview of the proposed technique for object retrieval from clutter with [PITH_FULL_IMAGE:figures/full_fig_p051_4_2.png]
Figure 4
Figure 4. Figure 4 [PITH_FULL_IMAGE:figures/full_fig_p052_4.png]
Figure 4
Figure 4. Figure 4 [PITH_FULL_IMAGE:figures/full_fig_p055_4.png]
Figure 4.3
Figure 4.3. Figure 4.3: Example of 4 consecutive pushes showing that DIPN can accurately predict push outcomes over a long horizon. We use purple arrows to illustrate push actions. The first and second columns are the predictions and ground truth (objects’ positions after executing the push…
Figure 4
Figure 4. Figure 4 [PITH_FULL_IMAGE:figures/full_fig_p056_4.png]
Figure 4
Figure 4. Figure 4 [PITH_FULL_IMAGE:figures/full_fig_p059_4.png]
Figure 4.4
Figure 4.4. Figure 4.4: 22 Test cases used in both simulation and real world experiments. The target [PITH_FULL_IMAGE:figures/full_fig_p060_4_4.png]
Figure 4
Figure 4. Figure 4 [PITH_FULL_IMAGE:figures/full_fig_p063_4.png]
Figure 4.5
Figure 4.5. Figure 4.5: Simulation results per test case for the [PITH_FULL_IMAGE:figures/full_fig_p064_4_5.png]
Figure 4
Figure 4. Figure 4: a [PITH_FULL_IMAGE:figures/full_fig_p064_4.png]
Figure 4.6
Figure 4.6. Figure 4.6: Simulation result per test case for the 22 harder problems ( [PITH_FULL_IMAGE:figures/full_fig_p065_4_6.png]
Figure 4.7
Figure 4.7. Figure 4.7: Real experiment results per test case for the [PITH_FULL_IMAGE:figures/full_fig_p065_4_7.png]
Figure 4
Figure 4. Figure 4 [PITH_FULL_IMAGE:figures/full_fig_p066_4.png]
Figure 4.8
Figure 4.8. Figure 4.8: Test scenario with soap boxes and masked (in purple) 3D printed vehicle. Two [PITH_FULL_IMAGE:figures/full_fig_p066_4_8.png]
Figure 5.1
Figure 5.1. Figure 5.1: (a) The hardware setup for object-retrieval-from-clutter includes a Universal [PITH_FULL_IMAGE:figures/full_fig_p069_5_1.png]
Figure 5
Figure 5. Figure 5 [PITH_FULL_IMAGE:figures/full_fig_p070_5.png]
Figure 5
Figure 5. Figure 5 [PITH_FULL_IMAGE:figures/full_fig_p073_5.png]
Figure 5.2
Figure 5.2. Figure 5.2: Sampled push actions. 5.3.2 Push Prediction Network (PPN) As previously mentioned, PPN learns to imitate MCTS. PPN is a deep neural network with ResNet-34 FPN [102], [110] as the backbone, where the P2 level of the FPN connects to the head. It takes a two-channel inp…
Figure 5
Figure 5. Figure 5 [PITH_FULL_IMAGE:figures/full_fig_p074_5.png]
Figure 5.3
Figure 5.3. Figure 5.3: The left two figures are the input to PPN. The first is a segmentation of objects; the second is the mask of the target object. The image on the right is the output from the PPN. We use Jet colormap to represent the reward value, where the value ranges from red (high…
Figure 5
Figure 5. Figure 5 [PITH_FULL_IMAGE:figures/full_fig_p076_5.png]
Figure 5
Figure 5. Figure 5 [PITH_FULL_IMAGE:figures/full_fig_p077_5.png]
Figure 5.4
Figure 5.4. Figure 5.4: An example of the guided MCTS with a budget of 10 iterations. State with [PITH_FULL_IMAGE:figures/full_fig_p077_5_4.png]
Figure 4
Figure 4. Figure 4 [PITH_FULL_IMAGE:figures/full_fig_p078_4.png]
Figure 5
Figure 5. Figure 5 [PITH_FULL_IMAGE:figures/full_fig_p080_5.png]
Figure 5.5
Figure 5.5. Figure 5.5: The average number (out of 5 trials) of action used to solve one case for 22 [PITH_FULL_IMAGE:figures/full_fig_p081_5_5.png]
Figure 5.6
Figure 5.6. Figure 5.6: The average time (of 5 trials) used to solve one case for 22 cases. [PITH_FULL_IMAGE:figures/full_fig_p081_5_6.png]
Figure 5.7
Figure 5.7. Figure 5.7: Different amounts of training data are used to train [PITH_FULL_IMAGE:figures/full_fig_p082_5_7.png]
Figure 5.8
Figure 5.8. Figure 5.8: Manually generated cases similar to [48] and Chapter 4. The target object is masked in purple. These cases are used also in simulation experiments as shown in [PITH_FULL_IMAGE:figures/full_fig_p082_5_8.png]
Figure 5
Figure 5. Figure 5 [PITH_FULL_IMAGE:figures/full_fig_p083_5.png]
Figure 5.9
Figure 5.9. Figure 5.9: The number of action and time used on solving six cases. The budget is up to [PITH_FULL_IMAGE:figures/full_fig_p083_5_9.png]
Figure 6.1
Figure 6.1. Figure 6.1: (a) The hardware setup includes a Universal Robots UR-5e with a Robotiq [PITH_FULL_IMAGE:figures/full_fig_p086_6_1.png]
Figure 6
Figure 6. Figure 6 [PITH_FULL_IMAGE:figures/full_fig_p088_6.png]
Figure 6
Figure 6. Figure 6 [PITH_FULL_IMAGE:figures/full_fig_p089_6.png]
Figure 6.2
Figure 6.2. Figure 6.2: Sampled push actions. Speeding up Action Sampling. Because the number of push action choices is un￾countably infinite, action sampling is necessary. We modified the action sampler from Chapter 4 and Chapter 5 with slight changes and a more efficient implementation. A…
Figure 6
Figure 6. Figure 6 [PITH_FULL_IMAGE:figures/full_fig_p091_6.png]
Figure 6.3
Figure 6.3. Figure 6.3: Examples of using the grasp classifier to produce probabilities to grasp the object at center (blue in this case). Here we used an RGB image for illustration purpose (input should be a depth image). network (GN) to evaluate the feasibility of grasping the target obje…
Figure 6
Figure 6. Figure 6 [PITH_FULL_IMAGE:figures/full_fig_p092_6.png]
Figure 6
Figure 6. Figure 6 [PITH_FULL_IMAGE:figures/full_fig_p094_6.png]
Figure 6.4
Figure 6.4. Figure 6.4: Steps in PMBS, our parallel MCTS with batched operation. [PITH_FULL_IMAGE:figures/full_fig_p094_6_4.png]
Figure 6
Figure 6. Figure 6 [PITH_FULL_IMAGE:figures/full_fig_p095_6.png]
Figure 6
Figure 6. Figure 6 [PITH_FULL_IMAGE:figures/full_fig_p096_6.png]
Figure 6
Figure 6. Figure 6 [PITH_FULL_IMAGE:figures/full_fig_p097_6.png]
Figure 6.5
Figure 6.5. Figure 6.5: 20 cases from Chapter 4 used in simulation experiments, where the target object has a blue mask. No object should exceed the boundary (red lines). Baseline. We use an optimized serial MCTS implementation as the baseline, where the number of environments used for MCTS…
Figure 6
Figure 6. Figure 6 [PITH_FULL_IMAGE:figures/full_fig_p098_6.png]
Figure 6
Figure 6. Figure 6 [PITH_FULL_IMAGE:figures/full_fig_p099_6.png]
Figure 6.6
Figure 6.6. Figure 6.6: The average number (over five independent trials) of actions per case needed for [PITH_FULL_IMAGE:figures/full_fig_p099_6_6.png]
Figure 6.7
Figure 6.7. Figure 6.7: The average time (over five independent trials) per case needed for solving the [PITH_FULL_IMAGE:figures/full_fig_p099_6_7.png]
Figure 6
Figure 6. Figure 6 [PITH_FULL_IMAGE:figures/full_fig_p100_6.png]
Figure 6.8
Figure 6.8. Figure 6.8: PMBS and serial MCTS evaluated with different time budgets. The reported [PITH_FULL_IMAGE:figures/full_fig_p101_6_8.png]
Figure 6
Figure 6. Figure 6 [PITH_FULL_IMAGE:figures/full_fig_p101_6.png]
Figure 6.9
Figure 6.9. Figure 6.9: The number of actions and time used for solving the six most challenging cases [PITH_FULL_IMAGE:figures/full_fig_p102_6_9.png]
Figure 7
Figure 7. Figure 7 [PITH_FULL_IMAGE:figures/full_fig_p103_7.png]
Figure 7.1
Figure 7.1. Figure 7.1: (a) Overview of system setup, a camera is mounted on the end-effector for [PITH_FULL_IMAGE:figures/full_fig_p104_7_1.png]
Figure 7
Figure 7. Figure 7 [PITH_FULL_IMAGE:figures/full_fig_p107_7.png]
Figure 7.2
Figure 7.2. Figure 7.2: Consider action sampling for labeled 3 to be manipulated using push (there are a total of four objects). The absence of sampled actions in the right region is attributed to obstructions posed by objects 0, 1, and 2, preventing the movement of object 3 to that area. c…
Figure 7.3
Figure 7.3. Figure 7.3: Example cases. The top row shows the start states and the bottom goal states. [PITH_FULL_IMAGE:figures/full_fig_p112_7_3.png]
Figure 7
Figure 7. Figure 7 [PITH_FULL_IMAGE:figures/full_fig_p113_7.png]
Figure 7.4
Figure 7.4. Figure 7.4: As an expanded illustration of [PITH_FULL_IMAGE:figures/full_fig_p113_7_4.png]
Figure 7
Figure 7. Figure 7 [PITH_FULL_IMAGE:figures/full_fig_p114_7.png]
Figure 7
Figure 7. Figure 7 [PITH_FULL_IMAGE:figures/full_fig_p115_7.png]
Figure 7.5
Figure 7.5. Figure 7.5: PMMR is evaluated with different time budgets. The reported values are averaged over 40 cases [PITH_FULL_IMAGE:figures/full_fig_p115_7_5.png]
Figure 7.6
Figure 7.6. Figure 7.6: The full set of objects used in our real-robot experiments. [PITH_FULL_IMAGE:figures/full_fig_p116_7_6.png]
Figure 7
Figure 7. Figure 7 [PITH_FULL_IMAGE:figures/full_fig_p117_7.png]
Figure 7.7
Figure 7.7. Figure 7.7: As an expanded illustration of [PITH_FULL_IMAGE:figures/full_fig_p117_7_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

153 extracted references · 66 canonical work pages

  1. [48]

    Efficient learning of goal-oriented push-grasping synergy in clutter,

    K. Xu, H. Yu, Q. Lai, Y. Wang, and R. Xiong, “Efficient learning of goal-oriented push-grasping synergy in clutter,” IEEE Robotics and Automation Letters, vol. 6, no. 4, pp. 6337–6344, 2021

    2021

  2. [1]

    Toward robotic manipulation,

    M. T. Mason, “Toward robotic manipulation,”Annual Review of Control, Robotics, and Autonomous Systems, vol. 1, no. 1, pp. 1–28, 2018

    2018

  3. [2]

    Robotized and automated warehouse systems: Review and recent developments,

    K. Azadeh, R. De Koster, and D. Roy, “Robotized and automated warehouse systems: Review and recent developments,”Transportation Science, vol. 53, no. 4, pp. 917– 945, 2019

    2019

  4. [3]

    Medical robotics and computer-integrated surgery,

    R. H. Taylor, A. Menciassi, G. Fichtinger, P. Fiorini, and P. Dario, “Medical robotics and computer-integrated surgery,” Springer handbook of robotics, pp. 1657–1684, 2016

    2016

  5. [4]

    Progress and prospects of the human–robot collaboration,

    A. Ajoudani, A. M. Zanchettin, S. Ivaldi, A. Albu-Sch ¨affer, K. Kosuge, and O. Khatib, “Progress and prospects of the human–robot collaboration,” Autonomous robots, vol. 42, pp. 957–975, 2018

    2018

  6. [5]

    Cooperative search and rescue with a team of mobile robots,

    J. S. Jennings, G. Whelan, and W. F. Evans, “Cooperative search and rescue with a team of mobile robots,” in 1997 8th International Conference on Advanced Robotics. Proceedings. ICAR’97, IEEE, 1997, pp. 193–200

    1997

  7. [6]

    Dex-net 3.0: Computing robust vacuum suction grasp targets in point clouds using a new analytic model and deep learning,

    J. Mahler, M. Matl, X. Liu, A. Li, D. Gealy, and K. Goldberg, “Dex-net 3.0: Computing robust vacuum suction grasp targets in point clouds using a new analytic model and deep learning,” in 2018 IEEE International Conference on robotics and automation (ICRA), IEEE, 2018, pp. 5620–5627

    2018

  8. [7]

    Fusebot: Rf-visual mechanical search,

    T. Boroushaki, L. Dodds, N. Naeem, and F. Adib, “Fusebot: Rf-visual mechanical search,” Robotics: Science and Systems 2022, 2022

    2022

Show all 153 references
  1. [8]

    Dealing with difficult instances of object rearrange- ment.,

    A. Krontiris and K. E. Bekris, “Dealing with difficult instances of object rearrange- ment.,” in Robotics: Science and Systems, vol. 1123, 2015

    2015

  2. [9]

    Dynamic grasp and trajectory planning for moving objects,

    N. Marturi, M. Kopicki, A. Rastegarpanah, et al., “Dynamic grasp and trajectory planning for moving objects,”Autonomous Robots, vol. 43, pp. 1241–1256, 2019

    2019

  3. [10]

    Learning dexterous in-hand manipulation,

    O. M. Andrychowicz, B. Baker, M. Chociej, et al., “Learning dexterous in-hand manipulation,” The International Journal of Robotics Research , vol. 39, no. 1, pp. 3–20, 2020

    2020

  4. [11]

    From caging to grasping,

    A. Rodriguez, M. T. Mason, and S. Ferry, “From caging to grasping,” The Interna- tional Journal of Robotics Research, vol. 31, no. 7, pp. 886–900, 2012. 111

    2012

  5. [12]

    Dipn: Deep interaction prediction network with application to clutter removal,

    B. Huang, S. D. Han, A. Boularias, and J. Yu, “Dipn: Deep interaction prediction network with application to clutter removal,” in2021 IEEE International Conference on Robotics and Automation (ICRA), IEEE, 2021, pp. 4694–4701

    2021

  6. [14]

    Toward optimal tabletop rearrangement with multiple manipulation primitives,

    B. Huang, X. Zhang, and J. Yu, “Toward optimal tabletop rearrangement with multiple manipulation primitives,” in 2024 IEEE International Conference on Robotics and Automation (ICRA), IEEE, 2024, pp. 10 860–10 866

    2024

  7. [15]

    Lgmcts: Language-guided monte- carlo tree search for executable semantic object rearrangement,

    H. Chang, K. Gao, K. Boyalakuntla, et al., “Lgmcts: Language-guided monte- carlo tree search for executable semantic object rearrangement,” in2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE, 2024, pp. 13 607–13 612

    2024

  8. [16]

    Visual foresight trees for ob- ject retrieval from clutter with nonprehensile rearrangement,

    B. Huang, S. D. Han, J. Yu, and A. Boularias, “Visual foresight trees for ob- ject retrieval from clutter with nonprehensile rearrangement,” IEEE Robotics and Automation Letters, vol. 7, no. 1, pp. 231–238, 2021

    2021

  9. [17]

    Toward fully automated metal recycling using computer vision and non-prehensile manipulation,

    S. D. Han, B. Huang, S. Ding, et al., “Toward fully automated metal recycling using computer vision and non-prehensile manipulation,” in 2021 IEEE 17th International Conference on Automation Science and Engineering (CASE), IEEE, 2021, pp. 891– 898

    2021

  10. [18]

    Fast high-quality tabletop rearrangement in bounded workspace,

    K. Gao, D. Lau, B. Huang, K. E. Bekris, and J. Yu, “Fast high-quality tabletop rearrangement in bounded workspace,” in2022 International Conference on Robotics and Automation (ICRA), IEEE, 2022, pp. 1961–1967

    2022

  11. [19]

    Interleaving monte carlo tree search and self-supervised learning for object retrieval in clutter,

    B. Huang, T. Guo, A. Boularias, and J. Yu, “Interleaving monte carlo tree search and self-supervised learning for object retrieval in clutter,” in 2022 International Conference on Robotics and Automation (ICRA), IEEE, 2022, pp. 625–632

    2022

  12. [20]

    Stackelberg strategic guidance for hetero- geneous robots collaboration,

    Y. Zhao, B. Huang, J. Yu, and Q. Zhu, “Stackelberg strategic guidance for hetero- geneous robots collaboration,” in 2022 International Conference on Robotics and Automation (ICRA), IEEE, 2022, pp. 4922–4928

    2022

  13. [21]

    Parallel monte carlo tree search with batched rigid-body simulations for speeding up long-horizon episodic robot planning,

    B. Huang, A. Boularias, and J. Yu, “Parallel monte carlo tree search with batched rigid-body simulations for speeding up long-horizon episodic robot planning,” in 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE, 2022, pp. 1153–1160. 112

    2022

  14. [22]

    Earl: Eye-on-hand reinforcement learner for dynamic grasping with active pose estimation,

    B. Huang, J. Yu, and S. Jain, “Earl: Eye-on-hand reinforcement learner for dynamic grasping with active pose estimation,” in 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE, 2023, pp. 2963–2970

    2023

  15. [23]

    Learning generalizable pivoting skills,

    X. Zhang, S. Jain, B. Huang, M. Tomizuka, and D. Romeres, “Learning generalizable pivoting skills,” in2023 IEEE International Conference on Robotics and Automation (ICRA), IEEE, 2023, pp. 5865–5871

    2023

  16. [24]

    Robotic grasping and contact: A review,

    A. Bicchi and V. Kumar, “Robotic grasping and contact: A review,” inProceedings 2000 ICRA. Millennium conference. IEEE international conference on robotics and automation. Symposia proceedings (Cat. No. 00CH37065), IEEE, vol. 1, 2000, pp. 348–353

    2000

  17. [25]

    A probabilistic data-driven model for planar pushing,

    M. Bauza and A. Rodriguez, “A probabilistic data-driven model for planar pushing,” in 2017 IEEE International Conference on Robotics and Automation (ICRA), IEEE, 2017, pp. 3008–3015

    2017

  18. [26]

    Towards robust product packing with a minimalistic end-effector,

    R. Shome, W. N. Tang, C. Song, et al., “Towards robust product packing with a minimalistic end-effector,” in 2019 International Conference on Robotics and Automation (ICRA), IEEE, 2019, pp. 9007–9013

    2019

  19. [27]

    Learning synergies between pushing and grasping with self-supervised deep reinforcement learning,

    A. Zeng, S. Song, S. Welker, J. Lee, A. Rodriguez, and T. Funkhouser, “Learning synergies between pushing and grasping with self-supervised deep reinforcement learning,” in 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE, 2018, pp. 4238–4245

    2018

  20. [28]

    Pre-grasp sliding manipulation of thin objects using soft, compliant, or underactuated hands,

    K. Hang, A. S. Morgan, and A. M. Dollar, “Pre-grasp sliding manipulation of thin objects using soft, compliant, or underactuated hands,” IEEE Robotics and Automation Letters, vol. 4, no. 2, pp. 662–669, 2019

    2019

  21. [29]

    Object search by manipulation,

    M. R. Dogar, M. C. Koval, A. Tallavajhula, and S. S. Srinivasa, “Object search by manipulation,” Autonomous Robots, vol. 36, pp. 153–167, 2014

    2014

  22. [30]

    Robust planar dynamic pivoting by regulating inertial and grip forces,

    Y. Hou, Z. Jia, A. M. Johnson, and M. T. Mason, “Robust planar dynamic pivoting by regulating inertial and grip forces,” in Algorithmic Foundations of Robotics XII: Proceedings of the Twelfth Workshop on the Algorithmic Foundations of Robotics, Springer, 2020, pp. 464–479

    2020

  23. [31]

    Manipulation of unknown objects via contact configuration regulation,

    N. Doshi, O. Taylor, and A. Rodriguez, “Manipulation of unknown objects via contact configuration regulation,” in 2022 International Conference on Robotics and Automation (ICRA), IEEE, 2022, pp. 2693–2699

    2022

  24. [32]

    Data-driven grasp synthesis—a survey,

    J. Bohg, A. Morales, T. Asfour, and D. Kragic, “Data-driven grasp synthesis—a survey,” IEEE Transactions on robotics, vol. 30, no. 2, pp. 289–309, 2013. 113

    2013

  25. [33]

    Learning a dictionary of prototypical grasp-predicting parts from grasping experience,

    R. Detry, C. H. Ek, M. Madry, and D. Kragic, “Learning a dictionary of prototypical grasp-predicting parts from grasping experience,” in 2013 IEEE International Conference on Robotics and Automation, IEEE, 2013, pp. 601–608

    2013

  26. [34]

    Deep learning for detecting robotic grasps,

    I. Lenz, H. Lee, and A. Saxena, “Deep learning for detecting robotic grasps,”The International Journal of Robotics Research, vol. 34, no. 4-5, pp. 705–724, 2015

    2015

  27. [35]

    Leveraging big data for grasp planning,

    D. Kappler, J. Bohg, and S. Schaal, “Leveraging big data for grasp planning,” in 2015 IEEE international conference on robotics and automation (ICRA) , IEEE, 2015, pp. 4304–4311

    2015

  28. [36]

    Learning 6-dof grasping interaction via deep 3d geometry-aware representations,

    X. Yan, J. Hsu, M. Khansari,et al., “Learning 6-dof grasping interaction via deep 3d geometry-aware representations,” in Proceedings of IEEE International Conference on Robotics and Automation (ICRA 2018), 2018

    2018

  29. [37]

    6-dof graspnet: Variational grasp generation for object manipulation,

    A. Mousavian, C. Eppner, and D. Fox, “6-dof graspnet: Variational grasp generation for object manipulation,” in Proceedings of the IEEE/CVF international conference on computer vision, 2019, pp. 2901–2910

    2019

  30. [38]

    Pointnetgpd: Detecting grasp configurations from point sets,

    H. Liang, X. Ma, S. Li, et al., “Pointnetgpd: Detecting grasp configurations from point sets,” in 2019 International Conference on Robotics and Automation (ICRA), IEEE, 2019, pp. 3629–3635

    2019

  31. [39]

    Using geometry to detect grasp poses in 3d point clouds,

    A. Ten Pas and R. Platt, “Using geometry to detect grasp poses in 3d point clouds,” Robotics Research: Volume 1, pp. 307–324, 2018

    2018

  32. [40]

    Supersizing self-supervision: Learning to grasp from 50k tries and 700 robot hours,

    L. Pinto and A. Gupta, “Supersizing self-supervision: Learning to grasp from 50k tries and 700 robot hours,” in 2016 IEEE international conference on robotics and automation (ICRA), IEEE, 2016, pp. 3406–3413

    2016

  33. [41]

    Learning deep policies for robot bin picking by simulating robust grasping sequences,

    J. Mahler and K. Goldberg, “Learning deep policies for robot bin picking by simulating robust grasping sequences,” in Conference on robot learning, PMLR, 2017, pp. 515–524

    2017

  34. [42]

    Graspnet-1billion: A large-scale bench- mark for general object grasping,

    H.-S. Fang, C. Wang, M. Gou, and C. Lu, “Graspnet-1billion: A large-scale bench- mark for general object grasping,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020, pp. 11 444–11 453

    2020

  35. [43]

    Efficient optimization for autonomous robotic manipulation of natural objects,

    A. Boularias, J. Bagnell, and A. Stentz, “Efficient optimization for autonomous robotic manipulation of natural objects,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 28, 2014

    2014

  36. [44]

    Catgrasp: Learning category-level task-relevant grasping in clutter from simulation,

    B. Wen, W. Lian, K. Bekris, and S. Schaal, “Catgrasp: Learning category-level task-relevant grasping in clutter from simulation,” in 2022 International Conference on Robotics and Automation (ICRA), IEEE, 2022, pp. 6401–6408. 114

    2022

  37. [45]

    Dex-net 2.0: Deep learning to plan robust grasps with synthetic point clouds and analytic grasp metrics,

    J. Mahler, J. Liang, S. Niyaz, et al., “Dex-net 2.0: Deep learning to plan robust grasps with synthetic point clouds and analytic grasp metrics,” arXiv preprint arXiv:1703.09312, 2017

    2017 arXiv

  38. [46]

    Learning ambidextrous robot grasping policies,

    J. Mahler, M. Matl, V. Satish,et al., “Learning ambidextrous robot grasping policies,” Science Robotics, vol. 4, no. 26, eaau4984, 2019

    2019

  39. [47]

    Deep reinforcement learning for robotic pushing and picking in cluttered environment,

    Y. Deng, X. Guo, Y. Wei,et al., “Deep reinforcement learning for robotic pushing and picking in cluttered environment,” in 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Ieee, 2019, pp. 619–626

    2019

  40. [49]

    Let’s push things forward: A survey on robot pushing,

    J. St¨ uber, C. Zito, and R. Stolkin, “Let’s push things forward: A survey on robot pushing,” Frontiers in Robotics and AI, vol. 7, p. 8, 2020

    2020

  41. [50]

    Estimating the friction parameters of pushed objects,

    K. M. Lynch, “Estimating the friction parameters of pushed objects,” inProceedings of 1993 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS’93), IEEE, vol. 1, 1993, pp. 186–193

    1993

  42. [51]

    Mechanics and planning of manipulator pushing operations,

    M. T. Mason, “Mechanics and planning of manipulator pushing operations,” The International Journal of Robotics Research, vol. 5, no. 3, pp. 53–71, 1986

    1986

  43. [52]

    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, vol. 15, no. 6, pp. 533–556, 1996

    1996

  44. [53]

    Dynamic nonprehensile manipulation: Controllabil- ity, planning, and experiments,

    K. M. Lynch and M. T. Mason, “Dynamic nonprehensile manipulation: Controllabil- ity, planning, and experiments,” The International Journal of Robotics Research, vol. 18, no. 1, pp. 64–92, 1999

    1999

  45. [54]

    Posing polygonal objects in the plane by pushing,

    S. Akella and M. T. Mason, “Posing polygonal objects in the plane by pushing,”The International Journal of Robotics Research, vol. 17, no. 1, pp. 70–88, 1998

    1998

  46. [55]

    On the scope of quasi-static pushing,

    M. T. Mason, “On the scope of quasi-static pushing,” inInternational Symposium on Robotics Research, 1986, 1986, pp. 229–233

    1986

  47. [56]

    Indentification of the center of friction from pushing an object by a mobile robot,

    T. Yoshikawa and M. Kurisu, “Indentification of the center of friction from pushing an object by a mobile robot,” in Proceedings IROS’91: IEEE/RSJ International Workshop on Intelligent Robots and Systems’ 91, IEEE, 1991, pp. 449–454. 115

    1991

  48. [57]

    Practical force-motion models for sliding manipulation,

    R. D. Howe and M. R. Cutkosky, “Practical force-motion models for sliding manipulation,” The International Journal of Robotics Research , vol. 15, no. 6, pp. 557–572, 1996

    1996

  49. [58]

    A convex polynomial force- motion model for planar sliding: Identification and application,

    J. Zhou, R. Paolini, J. A. Bagnell, and M. T. Mason, “A convex polynomial force- motion model for planar sliding: Identification and application,” in 2016 IEEE International Conference on Robotics and Automation (ICRA), IEEE, 2016, pp. 372– 377

    2016

  50. [59]

    A convex polynomial model for planar sliding mechanics: Theory, application, and experimental validation,

    J. Zhou, M. T. Mason, R. Paolini, and D. Bagnell, “A convex polynomial model for planar sliding mechanics: Theory, application, and experimental validation,” The International Journal of Robotics Research, vol. 37, no. 2-3, pp. 249–265, 2018

    2018

  51. [60]

    Pushing revisited: Differential flatness, trajectory planning, and stabilization,

    J. Zhou, Y. Hou, and M. T. Mason, “Pushing revisited: Differential flatness, trajectory planning, and stabilization,”The International Journal of Robotics Research, vol. 38, no. 12-13, pp. 1477–1489, 2019

    2019

  52. [61]

    A framework for push-grasping in clutter.,

    M. R. Dogar and S. S. Srinivasa, “A framework for push-grasping in clutter.,” in Robotics: Science and systems, vol. 2, 2011

    2011

  53. [62]

    Unsupervised learning for physical interaction through video prediction,

    C. Finn, I. Goodfellow, and S. Levine, “Unsupervised learning for physical interaction through video prediction,” in Advances in neural information processing systems, 2016, pp. 64–72

    2016

  54. [63]

    Se3-nets: Learning rigid body motion using deep neural networks,

    A. Byravan and D. Fox, “Se3-nets: Learning rigid body motion using deep neural networks,” in 2017 IEEE International Conference on Robotics and Automation (ICRA), IEEE, 2017, pp. 173–180

    2017

  55. [64]

    Visual interaction networks: Learning a physics simulator from video,

    N. Watters, D. Zoran, T. Weber, P. Battaglia, R. Pascanu, and A. Tacchetti, “Visual interaction networks: Learning a physics simulator from video,” in Advances in neural information processing systems, 2017, pp. 4539–4547

    2017

  56. [65]

    Learning contact-rich manipulation skills with guided policy search,

    L. Sergey, N. Wagener, and P. Abbeel, “Learning contact-rich manipulation skills with guided policy search,” in Proceedings of the 2015 IEEE International Conference on Robotics and Automation (ICRA), Seattle, WA, USA, 2015, pp. 26–30

    2015

  57. [66]

    End-to-end training of deep visuomotor policies,

    S. Levine, C. Finn, T. Darrell, and P. Abbeel, “End-to-end training of deep visuomotor policies,” Journal of Machine Learning Research, vol. 17, no. 39, pp. 1–40, 2016

    2016

  58. [67]

    Deep visual foresight for planning robot motion,

    C. Finn and S. Levine, “Deep visual foresight for planning robot motion,” in2017 IEEE International Conference on Robotics and Automation (ICRA), IEEE, 2017, pp. 2786–2793. 116

    2017

  59. [68]

    Deep predictive policy training using reinforcement learning,

    A. Ghadirzadeh, A. Maki, D. Kragic, and M. Bj ¨orkman, “Deep predictive policy training using reinforcement learning,” in 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE, 2017, pp. 2351–2358

    2017

  60. [69]

    Push-grasping with dexterous hands: Mechanics and a method,

    M. R. Dogar and S. S. Srinivasa, “Push-grasping with dexterous hands: Mechanics and a method,” in 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems, IEEE, 2010, pp. 2123–2130

    2010

  61. [70]

    Physics-based grasp planning through clutter.,

    M. R. Dogar, K. Hsiao, M. T. Ciocarlie, and S. S. Srinivasa, “Physics-based grasp planning through clutter.,” in Robotics: Science and systems, vol. 8, 2012, pp. 57–64

    2012

  62. [71]

    Pregrasp manipulation as trajectory optimization.,

    J. E. King, M. Klingensmith, C. M. Dellin,et al., “Pregrasp manipulation as trajectory optimization.,” in Robotics: Science and Systems, Berlin, 2013

    2013

  63. [72]

    Interactive singulation of objects from a pile,

    L. Chang, J. R. Smith, and D. Fox, “Interactive singulation of objects from a pile,” in 2012 IEEE International Conference on Robotics and Automation, IEEE, 2012, pp. 3875–3882

    2012

  64. [73]

    Learning to singulate objects using a push proposal network,

    A. Eitel, N. Hauff, and W. Burgard, “Learning to singulate objects using a push proposal network,” in Robotics Research: The 18th International Symposium ISRR, Springer, 2020, pp. 405–419

    2020

  65. [74]

    Linear push policies to increase grasp access for robot bin picking,

    M. Danielczuk, J. Mahler, C. Correa, and K. Goldberg, “Linear push policies to increase grasp access for robot bin picking,” in 2018 IEEE 14th international conference on automation science and engineering (CASE), IEEE, 2018, pp. 1249– 1256

    2018

  66. [75]

    Learning collabo- rative pushing and grasping policies in dense clutter,

    B. Tang, M. Corsaro, G. Konidaris, S. Nikolaidis, and S. Tellex, “Learning collabo- rative pushing and grasping policies in dense clutter,” in2021 IEEE International Conference on Robotics and Automation (ICRA), IEEE, 2021, pp. 6177–6184

    2021

  67. [76]

    Online planning for target object search in clutter under partial observability,

    Y. Xiao, S. Katt, A. ten Pas, S. Chen, and C. Amato, “Online planning for target object search in clutter under partial observability,” in2019 International Conference on Robotics and Automation (ICRA), IEEE, 2019, pp. 8241–8247

    2019

  68. [77]

    Mechanical search: Multi-step retrieval of a target object occluded by clutter,

    M. Danielczuk, A. Kurenkov, A. Balakrishna, et al., “Mechanical search: Multi-step retrieval of a target object occluded by clutter,” in2019 International Conference on Robotics and Automation (ICRA), IEEE, 2019, pp. 1614–1621

    2019

  69. [78]

    Visuomotor mechanical search: Learning to retrieve target objects in clutter,

    A. Kurenkov, J. Taglic, R. Kulkarni,et al., “Visuomotor mechanical search: Learning to retrieve target objects in clutter,” in IEEE/RSJ Int. Conference. on Intelligent Robots and Systems (IROS), 2020

    2020

  70. [79]

    Multi-object rearrangement with monte carlo tree search: A case study on planar nonprehensile sorting,

    H. Song, J. A. Haustein, W. Yuan,et al., “Multi-object rearrangement with monte carlo tree search: A case study on planar nonprehensile sorting,” in 2020 IEEE/RSJ 117 international conference on intelligent robots and systems (IROS) , IEEE, 2020, pp. 9433–9440

    2020

  71. [80]

    High-quality tabletop rearrangement with overhand grasps: Hardness results and fast methods,

    S. D. Han, N. M. Stiffler, A. Krontiris, K. E. Bekris, and J. Yu, “High-quality tabletop rearrangement with overhand grasps: Hardness results and fast methods,” in Robotics: Sciences and Systems, 2017

    2017

  72. [81]

    Complexity results and fast methods for optimal tabletop rearrangement with overhand grasps,

    S. D. Han, N. M. Stiffler, A. Krontiris, K. E. Bekris, and J. Yu, “Complexity results and fast methods for optimal tabletop rearrangement with overhand grasps,” The International Journal of Robotics Research, vol. 37, no. 13-14, pp. 1775–1795, 2018

    2018

  73. [82]

    On minimizing the number of running buffers for tabletop rearrangement,

    K. Gao, S. W. Feng, and J. Yu, “On minimizing the number of running buffers for tabletop rearrangement,” in Robotics: Sciences and Systems, 2021

    2021

  74. [83]

    Rearrangement on lattices with swaps: Optimality structures and efficient algorithms,

    J. Yu, “Rearrangement on lattices with swaps: Optimality structures and efficient algorithms,” in Robotics: Sciences and Systems, 2021

    2021

  75. [84]

    Rearrangement on lattices with pick-n-swaps: Optimality structures and efficient algorithms,

    J. Yu, “Rearrangement on lattices with pick-n-swaps: Optimality structures and efficient algorithms,” The International Journal of Robotics Research, vol. 42, no. 10, pp. 957–973, 2023

    2023

  76. [85]

    Minimizing running buffers for tabletop ob- ject rearrangement: Complexity, fast algorithms, and applications,

    K. Gao, S. W. Feng, B. Huang, and J. Yu, “Minimizing running buffers for tabletop ob- ject rearrangement: Complexity, fast algorithms, and applications,”The International Journal of Robotics Research, vol. 42, no. 10, pp. 755–776, 2023

    2023

  77. [86]

    Selective object rearrangement in clutter,

    B. Tang and G. S. Sukhatme, “Selective object rearrangement in clutter,” inConfer- ence on Robot Learning, PMLR, 2023, pp. 1001–1010

    2023

  78. [87]

    Coordination of two robotic manipulators for object retrieval in clutter,

    J. Ahn, C. Kim, and C. Nam, “Coordination of two robotic manipulators for object retrieval in clutter,” in 2022 International Conference on Robotics and Automation (ICRA), IEEE, 2022, pp. 1039–1045

    2022

  79. [88]

    Randomized physics-based motion planning for grasping in cluttered and uncertain environments,

    M. Moll, L. Kavraki, J. Rosell,et al., “Randomized physics-based motion planning for grasping in cluttered and uncertain environments,” IEEE Robotics and Automation Letters, vol. 3, no. 2, pp. 712–719, 2017

    2017

  80. [89]

    Efficient and high-quality prehensile rear- rangement in cluttered and confined spaces,

    R. Wang, Y. Miao, and K. E. Bekris, “Efficient and high-quality prehensile rear- rangement in cluttered and confined spaces,” in 2022 International Conference on Robotics and Automation (ICRA), IEEE, 2022, pp. 1968–1975

    2022

  81. [90]

    Hierarchical task and motion planning in the now,

    L. P. Kaelbling and T. Lozano-P´erez, “Hierarchical task and motion planning in the now,” in 2011 IEEE International Conference on Robotics and Automation, IEEE, 2011, pp. 1470–1477. 118

    2011

  82. [91]

    Combined task and motion planning through an extensible planner-independent interface layer,

    S. Srivastava, E. Fang, L. Riano, R. Chitnis, S. Russell, and P. Abbeel, “Combined task and motion planning through an extensible planner-independent interface layer,” in 2014 IEEE international conference on robotics and automation (ICRA), IEEE, 2014, pp. 639–646

    2014

  83. [92]

    Logic-geometric programming: An optimization-based approach to combined task and motion planning.,

    M. Toussaint, “Logic-geometric programming: An optimization-based approach to combined task and motion planning.,” in IJCAI, 2015, pp. 1930–1936

    2015

  84. [93]

    Incremental task and motion planning: A constraint-based approach.,

    N. T. Dantam, Z. K. Kingston, S. Chaudhuri, and L. E. Kavraki, “Incremental task and motion planning: A constraint-based approach.,” in Robotics: Science and systems, Ann Arbor, MI, USA, vol. 12, 2016, p. 00 052

    2016

  85. [94]

    Pddlstream: Integrating symbolic planners and blackbox samplers via optimistic adaptive planning,

    C. R. Garrett, T. Lozano-P ´erez, and L. P. Kaelbling, “Pddlstream: Integrating symbolic planners and blackbox samplers via optimistic adaptive planning,” in Proceedings of the international conference on automated planning and scheduling, vol. 30, 2020, pp. 440–448

    2020

  86. [95]

    Object-centric task and motion planning in dynamic environments,

    T. Migimatsu and J. Bohg, “Object-centric task and motion planning in dynamic environments,” IEEE Robotics and Automation Letters, vol. 5, no. 2, pp. 844–851, 2020

    2020

  87. [96]

    Guided search for task and motion plans using learned heuristics,

    R. Chitnis, D. Hadfield-Menell, A. Gupta, et al., “Guided search for task and motion plans using learned heuristics,” in 2016 IEEE International Conference on Robotics and Automation (ICRA), IEEE, 2016, pp. 447–454

    2016

  88. [97]

    Learning to guide task and motion planning using score-space representation,

    B. Kim, Z. Wang, L. P. Kaelbling, and T. Lozano-P´erez, “Learning to guide task and motion planning using score-space representation,” The International Journal of Robotics Research, vol. 38, no. 7, pp. 793–812, 2019

    2019

  89. [98]

    Deep visual reasoning: Learning to predict action sequences for task and motion planning from an initial scene image,

    D. Driess, J.-S. Ha, and M. Toussaint, “Deep visual reasoning: Learning to predict action sequences for task and motion planning from an initial scene image,”arXiv preprint arXiv:2006.05398, 2020

    2006 arXiv

  90. [99]

    Learning feasibility for task and motion planning in tabletop environments,

    A. M. Wells, N. T. Dantam, A. Shrivastava, and L. E. Kavraki, “Learning feasibility for task and motion planning in tabletop environments,” IEEE robotics and automation letters, vol. 4(2), pp. 1255–1262, 2019

    2019

  91. [100]

    Learning goal-oriented non-prehensile pushing in cluttered scenes,

    N. Dengler, D. Großklaus, and M. Bennewitz, “Learning goal-oriented non-prehensile pushing in cluttered scenes,” in 2022 IEEE/RSJ International Conference on Intelli- gent Robots and Systems (IROS), IEEE, 2022, pp. 1116–1122

    2022

  92. [101]

    Mask r-cnn,

    K. He, G. Gkioxari, P. Doll´ar, and R. Girshick, “Mask r-cnn,” in Proceedings of the IEEE international conference on computer vision, 2017, pp. 2961–2969. 119

    2017

  93. [102]

    Feature pyramid networks for object detection,

    T. Lin, P. Doll´ar, R. B. Girshick, K. He, B. Hariharan, and S. J. Belongie, “Feature pyramid networks for object detection,”CoRR, vol. abs/1612.03144, 2016. arXiv: 1612.03144

    2016 arXiv

  94. [103]

    Learning to see before learning to act: Visual pre-training for manipulation,

    L. Yen-Chen, A. Zeng, S. Song, P. Isola, and T. -Y. Lin, “Learning to see before learning to act: Visual pre-training for manipulation,” in 2020 IEEE International Conference on Robotics and Automation (ICRA), IEEE, 2020, pp. 7286–7293

    2020

  95. [104]

    V-rep: A versatile and scalable robot simulation framework,

    E. Rohmer, S. P. Singh, and M. Freese, “V-rep: A versatile and scalable robot simulation framework,” in2013 IEEE/RSJ International Conference on Intelligent Robots and Systems, IEEE, 2013, pp. 1321–1326

    2013

  96. [105]

    Dream to control: Learning behaviors by latent imagination,

    D. Hafner, T. Lillicrap, J. Ba, and M. Norouzi, “Dream to control: Learning behaviors by latent imagination,” in International Conference on Learning Representations, 2020

    2020

  97. [106]

    Visual foresight: Model-based deep reinforcement learning for vision-based robotic control,

    F. Ebert, C. Finn, S. Dasari, A. Xie, A. X. Lee, and S. Levine, “Visual foresight: Model-based deep reinforcement learning for vision-based robotic control,” CoRR, vol. abs/1812.00568, 2018. arXiv: 1812.00568

    2018 arXiv

  98. [107]

    Randomized physics-based motion planning for grasping in cluttered and uncertain environments,

    Muhayyuddin, M. Moll, L. Kavraki, and J. Rosell, “Randomized physics-based motion planning for grasping in cluttered and uncertain environments,” IEEE Robotics and Automation Letters, vol. 3, no. 2, pp. 712–719, Apr. 2018

    2018

  99. [108]

    Human-level control through deep reinforcement learning,

    V. Mnih, K. Kavukcuoglu, D. Silver, et al., “Human-level control through deep reinforcement learning,” nature, vol. 518, no. 7540, pp. 529–533, 2015

    2015

  100. [109]

    Hindsight experience replay,

    M. Andrychowicz, F. Wolski, A. Ray,et al., “Hindsight experience replay,”Advances in neural information processing systems, vol. 30, 2017

    2017

  101. [110]

    Deep residual learning for image recognition,

    K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 770–778

    2016

  102. [111]

    A survey of monte carlo tree search methods,

    C. B. Browne, E. Powley, D. Whitehouse,et al., “A survey of monte carlo tree search methods,” IEEE Transactions on Computational Intelligence and AI in Games, vol. 4, no. 1, pp. 1–43, 2012

    2012

  103. [112]

    Sgdr: Stochastic gradient descent with warm restarts,

    I. Loshchilov and F. Hutter, “Sgdr: Stochastic gradient descent with warm restarts,” arXiv preprint arXiv:1608.03983, 2016

    2016 arXiv

  104. [113]

    Hindsight experience replay,

    M. Andrychowicz, F. Wolski, A. Ray, et al., “Hindsight experience replay,” in Advances in Neural Information Processing Systems, I. Guyon, U. V. Luxburg, S. Bengio, et al., Eds., vol. 30, 2017. 120

    2017

  105. [114]

    Coumans and Y

    E. Coumans and Y. Bai, Pybullet, a python module for physics simulation for games, robotics and machine learning, http://pybullet.org, 2016–2019

    2016

  106. [115]

    Mastering atari, go, chess and shogi by planning with a learned model,

    J. Schrittwieser, I. Antonoglou, T. Hubert, et al., “Mastering atari, go, chess and shogi by planning with a learned model,” Nature, vol. 588, no. 7839, pp. 604–609, 2020

    2020

  107. [116]

    Kahneman, Thinking, fast and slow

    D. Kahneman, Thinking, fast and slow. Macmillan, 2011

    2011

  108. [117]

    The consciousness prior,

    Y. Bengio, “The consciousness prior,” arXiv preprint arXiv:1709.08568, 2017

    2017 arXiv

  109. [118]

    R. S. Sutton and A. G. Barto, Reinforcement learning: An introduction. MIT press, 2018

    2018

  110. [119]

    A general reinforcement learning algorithm that masters chess, shogi, and go through self-play,

    D. Silver, T. Hubert, J. Schrittwieser, et al., “A general reinforcement learning algorithm that masters chess, shogi, and go through self-play,” Science, vol. 362, no. 6419, pp. 1140–1144, 2018

    2018

  111. [120]

    Multi-object rearrangement with monte carlo tree search: A case study on planar nonprehensile sorting,

    H. Song, J. A. Haustein, W. Yuan, et al. , “Multi-object rearrangement with monte carlo tree search: A case study on planar nonprehensile sorting,” CoRR, vol. abs/1912.07024, 2019. arXiv: 1912.07024

    1912 arXiv

  112. [121]

    Regularizing trajectory optimization with denoising autoencoders,

    R. Boney, N. Di Palo, M. Berglund, et al., “Regularizing trajectory optimization with denoising autoencoders,” Advances in Neural Information Processing Systems, vol. 32, pp. 2859–2869, 2019

    2019

  113. [122]

    Efficient selectivity and backup operators in monte-carlo tree search,

    R. Coulom, “Efficient selectivity and backup operators in monte-carlo tree search,” in International conference on computers and games, Springer, 2006, pp. 72–83

    2006

  114. [123]

    Pytorch: An imperative style, high-performance deep learning library,

    A. Paszke, S. Gross, F. Massa,et al., “Pytorch: An imperative style, high-performance deep learning library,” in Advances in Neural Information Processing Systems 32, H. Wallach, H. Larochelle, A. Beygelzimer, F. d’ Alch´e-Buc, E. Fox, and R. Garnett, Eds., Curran Associates, ...

    2019

  115. [124]

    Combining q-learning and search with amortized value estimates,

    J. B. Hamrick, V. Bapst, A. Sanchez-Gonzalez,et al., “Combining q-learning and search with amortized value estimates,” in International Conference on Learning Representations ICLR, 2019

    2019

  116. [125]

    Coumans and Y

    E. Coumans and Y. Bai, Pybullet, a python module for physics simulation for games, robotics and machine learning, http://pybullet.org, 2016–2021

    2016

  117. [126]

    Se (3)-tracknet: Data-driven 6d pose tracking by calibrating image residuals in synthetic domains,

    B. Wen, C. Mitash, B. Ren, and K. E. Bekris, “Se (3)-tracknet: Data-driven 6d pose tracking by calibrating image residuals in synthetic domains,” in 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE, 2020, pp. 10 367–10 373. 121

    2020

  118. [127]

    Bundletrack: 6d pose tracking for novel objects without instance or category-level 3d models,

    B. Wen and K. Bekris, “Bundletrack: 6d pose tracking for novel objects without instance or category-level 3d models,” in 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE, 2021, pp. 8067–8074

    2021

  119. [128]

    Scene-level pose estimation for multiple instances of densely packed objects,

    C. Mitash, B. Wen, K. Bekris, and A. Boularias, “Scene-level pose estimation for multiple instances of densely packed objects,” in Conference on Robot Learning, PMLR, 2020, pp. 1133–1145

    2020

  120. [129]

    Learning agile robotic locomotion skills by imitating animals,

    X. B. Peng, E. Coumans, T. Zhang, T.-W. E. Lee, J. Tan, and S. Levine, “Learning agile robotic locomotion skills by imitating animals,” in Robotics: Science and Systems, Jul. 2020

    2020

  121. [130]

    Learning agile and dynamic motor skills for legged robots,

    J. Hwangbo, J. Lee, A. Dosovitskiy,et al., “Learning agile and dynamic motor skills for legged robots,” Science Robotics, vol. 4, no. 26, eaau5872, 2019

    2019

  122. [131]

    RMA: Rapid Motor Adaptation for Legged Robots,

    A. Kumar, Z. Fu, D. Pathak, and J. Malik, “RMA: Rapid Motor Adaptation for Legged Robots,” in Proceedings of Robotics: Science and Systems , Virtual, Jul. 2021

    2021

  123. [132]

    Imagenet classification with deep convolutional neural networks,

    A. Krizhevsky, I. Sutskever, and G. E. Hinton, “Imagenet classification with deep convolutional neural networks,”Advances in neural information processing systems, vol. 25, 2012

    2012

  124. [133]

    Playing atari with deep reinforcement learning,

    V. Mnih, K. Kavukcuoglu, D. Silver,et al., “Playing atari with deep reinforcement learning,” arXiv preprint arXiv:1312.5602, 2013

    2013 arXiv

  125. [134]

    MuJoCo: A physics engine for model-based control,

    E. Todorov, T. Erez, and Y. Tassa, “MuJoCo: A physics engine for model-based control,” in 2012 IEEE/RSJ international conference on intelligent robots and systems, IEEE, 2012, pp. 5026–5033

    2012

  126. [135]

    Isaac gym: High performance gpu-based physics simulation for robot learning,

    V. Makoviychuk, L. Wawrzyniak, Y. Guo,et al., “Isaac gym: High performance gpu-based physics simulation for robot learning,” arXiv preprint arXiv:2108.10470, 2021

    2021 arXiv

  127. [136]

    C. D. Freeman, E. Frey, A. Raichuk, S. Girgin, I. Mordatch, and O. Bachem, Brax - a differentiable physics engine for large scale rigid body simulation, version 0.0.10, 2021

    2021

  128. [137]

    Mastering chess and shogi by self-play with a general reinforcement learning algorithm,

    D. Silver, T. Hubert, J. Schrittwieser, et al., “Mastering chess and shogi by self-play with a general reinforcement learning algorithm,” arXiv preprint arXiv:1712.01815, 2017

    2017 arXiv

  129. [138]

    Finite-time analysis of the multiarmed bandit problem,

    P. Auer, N. Cesa-Bianchi, and P. Fischer, “Finite-time analysis of the multiarmed bandit problem,” Machine learning, vol. 47, no. 2, pp. 235–256, 2002. 122

    2002

  130. [139]

    Bandit based monte-carlo planning,

    L. Kocsis and C. Szepesv ´ari, “Bandit based monte-carlo planning,” in European conference on machine learning, Springer, 2006, pp. 282–293

    2006

  131. [140]

    Efficientnet: Rethinking model scaling for convolutional neural networks,

    M. Tan and Q. Le, “Efficientnet: Rethinking model scaling for convolutional neural networks,” inInternational conference on machine learning, PMLR, 2019, pp. 6105– 6114

    2019

  132. [141]

    Parallel monte-carlo tree search,

    G. M.-B. Chaslot, M. H. Winands, and H. Herik, “Parallel monte-carlo tree search,” in International Conference on Computers and Games, Springer, 2008, pp. 60–71

    2008

  133. [142]

    Watch the unobserved: A simple approach to parallelizing monte carlo tree search,

    A. Liu, J. Chen, M. Yu, Y. Zhai, X. Zhou, and J. Liu, “Watch the unobserved: A simple approach to parallelizing monte carlo tree search,” inInternational Conference on Learning Representations, 2020

    2020

  134. [143]

    Practical massively parallel monte-carlo tree search applied to molecular design,

    X. Yang, T. Aasawat, and K. Yoshizoe, “Practical massively parallel monte-carlo tree search applied to molecular design,” in International Conference on Learning Representations, 2021

    2021

  135. [144]

    Analytic inverse kinematics for the universal robots ur-5/ur-10 arms,

    K. P. Hawkins, “Analytic inverse kinematics for the universal robots ur-5/ur-10 arms,” Georgia Institute of Technology, Tech. Rep., 2013

    2013

  136. [145]

    Team rubot’s experiences and lessons from the ariac,

    S. W. Feng, T. Guo, K. E. Bekris, and J. Yu, “Team rubot’s experiences and lessons from the ariac,”Robotics and computer-integrated manufacturing, vol. 70, p. 102 126, 2021

    2021

  137. [146]

    Mastering the game of go with deep neural networks and tree search,

    D. Silver, A. Huang, C. J. Maddison, et al., “Mastering the game of go with deep neural networks and tree search,” nature, vol. 529, no. 7587, pp. 484–489, 2016

    2016

  138. [147]

    Monte-carlo tree search for efficient visually guided rearrangement planning,

    Y. Labb´e, S. Zagoruyko, I. Kalevatykh, et al., “Monte-carlo tree search for efficient visually guided rearrangement planning,” IEEE Robotics and Automation Letters, vol. 5, no. 2, pp. 3715–3722, 2020

    2020

  139. [148]

    Effectively rearranging heterogeneous objects on cluttered tabletops,

    K. Gao, J. Yu, T. S. Punjabi, and J. Yu, “Effectively rearranging heterogeneous objects on cluttered tabletops,” in 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE, 2023, pp. 2057–2064

    2023

  140. [149]

    Rrt-connect: An efficient approach to single-query path planning,

    J. Kuffner and S. LaValle, “Rrt-connect: An efficient approach to single-query path planning,” in Proceedings 2000 ICRA. Millennium Conference. IEEE Inter- national Conference on Robotics and Automation. Symposia Proceedings (Cat. No.00CH37065), vol. 2, 2000, 995–1001 vol.2

    2000

  141. [150]

    A randomized approach to robot path planning based on lazy evaluation,

    R. Bohlin and L. E. Kavraki, “A randomized approach to robot path planning based on lazy evaluation,” COMBINATORIAL OPTIMIZATION-DORDRECHT-, vol. 9, no. 1, pp. 221–249, 2001. 123

    2001

  142. [151]

    Hierarchical policy with deep- reinforcement learning for nonprehensile multiobject rearrangement,

    F. Bai, F. Meng, J. Liu, J. Wang, and M. Q.-H. Meng, “Hierarchical policy with deep- reinforcement learning for nonprehensile multiobject rearrangement,”Biomimetic Intelligence and Robotics, vol. 2, no. 3, p. 100 047, 2022

    2022

  143. [152]

    Pybullet, a python module for physics simulation for games, robotics and machine learning,

    E. Coumans and Y. Bai, “Pybullet, a python module for physics simulation for games, robotics and machine learning,” 2016

    2016

  144. [153]

    Segment anything,

    A. Kirillov, E. Mintun, N. Ravi, et al., “Segment anything,” in Proceedings of the IEEE/CVF international conference on computer vision, 2023, pp. 4015–4026

    2023

  145. [154]

    The OpenCV Library,

    G. Bradski, “The OpenCV Library,” Dr. Dobb’s Journal of Software Tools, 2000

    2000

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.