CLONE: Closed-Loop Whole-Body Humanoid Teleoperation for Long-Horizon Tasks

Jieming Cui; Siyuan Huang; Tengyu Liu; Wei Liang; Yixin Zhu; Yixuan Li; Yutang Lin

REVIEW 4 major objections 6 minor 9 cited by

CLONE: Closed-Loop Whole-Body Humanoid Teleoperation for Long-Horizon Tasks

T0 review · 4 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read A closed-loop teleoperation system keeps a humanoid within centimeters of an operator over long trajectories using only head and hand tracking from an MR headset.

desk verdict Solid closed-loop teleoperation integration, but the 5.1 cm drift number is measured against the same AVP reference the loop chases, so the global-accuracy claim outruns the evidence. read the letter →

arxiv 2506.08931 v2 pith:XJLPL5FD submitted 2025-06-10 cs.RO

Yixuan Li , Yutang Lin , Jieming Cui , Tengyu Liu , Wei Liang , Yixin Zhu , Siyuan Huang This is my paper · ORCID

classification cs.RO

keywords humanoidteleoperationwhole-bodycontrolmixture-of-expertsclosed-looperrorcorrectionLiDAR-inertialodometrylong-horizontasksloco-manipulationmotiondataset

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

CLONE is a humanoid teleoperation system that claims to solve two problems that have limited prior work: whole-body coordination and unbounded positional drift. It uses only the poses of both hands (position and orientation) and the position of the head from a commercial mixed-reality headset, and it learns a single Mixture-of-Experts policy that turns those sparse signals into joint commands for the whole body, including locomotion and squatting. Drift is addressed by closing the loop: LiDAR-inertial odometry gives the robot's global position, the headset gives the operator's global position, and the policy is trained to minimize the difference between the two. In real-world tests on a physical humanoid, the paper reports a mean straight-path tracking error of 5.1 cm up to 8.9 m, a maximum deviation of 12.0 cm, a curved-path mean error of 20 cm with 2 degrees of rotational drift over a 10 m trajectory with two 90-degree turns, and a 15 m mixed sequence that returns near its start. If these results hold, long-horizon teleoperation no longer needs full-body motion capture or open-loop control shortcuts.

What carries the argument

The load-bearing mechanism is a closed-loop global-position error signal: the robot's position $p \in \mathbb{R}^3$ comes from LiDAR-inertial odometry at 10 Hz, the operator's position $\hat{p} \in \mathbb{R}^3$ comes from the MR headset's odometry, and the policy's task observation includes the difference $\hat{p} - p$, so actions are chosen to reduce it. Around that signal, the student policy is a Mixture-of-Experts network with $L=3$ layers and $N=4$ experts per layer, taking the top $k=2$ experts by routing weight, with a balancing loss that keeps expert usage uniform; it is trained by distillation from a teacher policy and regularized by an adversarial motion prior that shapes lower-body behavior. A velocity-dependent stochastic noise model on the head position is added during training so the policy learns to tolerate odometry error that grows with movement speed.

What would settle it

Repeat the straight-path experiment while independently tracking both the operator and the robot with an external motion-capture or surveyed ground truth; if the robot's absolute endpoint error grows with distance while the operator-relative error stays small, the bounded drift is real, whereas if both errors grow together, the closed loop has only synchronized the two odometry drifts.

Watch

Extended reading notes

Core claim

On its own terms, the paper establishes that a teleoperation policy can preserve accurate global positioning over long trajectories while keeping whole-body coordination, using only sparse upper-body references. The core discovery is that including the closed-loop difference between the humanoid's LiDAR-estimated global position and the operator's headset-estimated global position in the policy observation changes the failure mode of teleoperation: instead of stepwise tracking error accumulating, the robot continuously corrects toward the operator, and the trained policy turns that correction into natural whole-body motion. The same policy, through routed expert specialization, reproduces standing, walking, turning, side-stepping, squatting, and jumping, with experts splitting along posture lines and all experts engaging during dynamic motions. The paper further claims that its curated CLONED dataset, though far smaller than prior motion-capture collections, outperforms the larger dataset on crouch and squat-to-pick tasks, and that the MoE architecture beats a monolithic network on local motion metrics and hand orientation tracking.

Load-bearing premise

The closed-loop controller treats the operator's global position, as estimated by the MR headset's motion tracking, as the true reference the robot should match, so any drift in that headset estimate becomes robot error.

Editorial extensions

If this is right

In the tested distance range, tracking error stops growing with distance: statistical tests across 3 m, 6 m, and 8.9 m show no significant degradation, so errors no longer compound per step.
Whole-body coordination, including squatting and standing from a squat, can be generated from sparse input alone, with lower-body behavior shaped by the adversarial motion prior rather than explicit lower-body commands.
A single policy can cover diverse skills without mode collapse when the student network uses routed experts rather than a monolithic network.
Dataset composition, including hand orientations and transition sequences, can matter more than raw dataset size for reproducing manipulation-relevant postures.
LiDAR-inertial odometry running at 10 Hz is sufficient to bound position error when the policy runs at 50 Hz and consumes the latest odometry reading.

Reading between the lines

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

The same closed-loop recipe transfers to other leader-follower pairs with independent odometry, so a different robot or a simulated avatar could inherit the drift bound without re-deriving the correction logic.
Performance should degrade in environments where the headset's visual-inertial odometry degrades, such as low light, sparse texture, or fast head motion, because the error signal itself becomes noisy.
If the operator reference drifts, the system converts unbounded robot drift into bounded operator-relative drift, so the headline error numbers should be read as error with respect to the operator, not with respect to the world.
The data-efficiency result suggests a cost-effective recipe for other whole-body control tasks: a few hundred curated, motion-edited sequences may beat much larger unfiltered collections.

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.

Desk Editor's Note

read the letter

CLONE is a solid systems integration paper, not a breakthrough. What is genuinely new is the package: an MoE student policy, LiDAR odometry feedback folded into the task observation, and a hand-orientation-augmented dataset, all running on a real Unitree G1 with commodity MR input. That combination is practically useful for humanoid teleoperation and data collection, and the real-robot demos—squatting, jumping, the 15 m mixed navigation—are credible evidence that the closed loop works at a systems level. The ablations against an MLP student and OmniH2O-style training data are meaningful, and the expert-activation analysis gives a real look at how the MoE specializes. The authors also cite the prior MoE and legged-odometry literature instead of hiding it. Credit where due: this is a working system, and the integration is nontrivial.

The soft spots are concentrated in the headline quantitative claim. The 5.1 cm mean error over 8.9 m is the discrepancy between the robot's LiDAR-estimated position and the operator's Apple Vision Pro position. The policy is explicitly trained to minimize exactly that difference, and the reward table confirms it. So the number measures how well the robot tracks AVP's own estimate, not how accurately the robot knows its global position. The paper never characterizes AVP odometry drift over these trajectories, and the frame alignment between AVP and LiDAR is described only as 'fixing the initial positions.' If AVP drifts slowly, the robot chases a moving target, and the reported error can stay small while both estimates are wrong. That gap is load-bearing for the 'unprecedented fidelity' and 'minimal positional drift' language in the abstract.

The other weaknesses are more moderate. There is no same-protocol open-loop or prior-system baseline, so the improvement claim over open-loop teleoperation is asserted rather than demonstrated. Table 1 is computed on CLONED, the training distribution, so those numbers measure in-distribution tracking; the stance-height experiments are edited from the same data, so generalization evidence is limited. No code, data, or trained models are released, which hurts reproducibility for a paper that is half dataset and architecture. There is also a leftover Chinese editing note in the Figure 2 caption that should be cleaned before any archival version.

Overall, the engineering story is plausible and the real-robot feasibility is credible, but the central global-accuracy number needs an independent ground-truth baseline before it supports the abstract's claims. This paper deserves a serious referee: the system is worth engaging with, and a good review could force the missing measurement and make the paper much stronger. Send it to review, with the clear instruction that the AVP-reference problem must be handled head-on.

Referee Report

4 major / 6 minor

Summary. The paper presents CLONE, a teleoperation system for a Unitree G1 humanoid that uses only head and hand tracking from an Apple Vision Pro headset. The system trains a teacher policy in simulation with privileged state information and distills it into a student MoE policy that observes robot state history and task observations, including the difference between the robot's LiDAR odometry position and the operator's MR odometry position. The authors report real-world straight-path tracking with a mean error of 5.1 cm over 8.9 m, curved-path tracking with 20 cm mean error over 10 m, whole-body skill demonstrations, and simulation comparisons against an MLP baseline and a dataset ablation. The central claim is that closed-loop error correction prevents drift accumulation in long-horizon whole-body teleoperation while preserving coordination.

Significance. If the central drift claim is properly supported, CLONE would be a practically valuable contribution: it uses minimal commercial hardware, demonstrates real-world whole-body skills including squatting and jumping on a physical humanoid, and the closed-loop formulation directly addresses a known weakness of open-loop teleoperation. The MoE architecture and the CLONED dataset with hand-orientation augmentation are plausible contributions. However, the headline quantitative claim currently depends on an uncharacterized external reference and lacks a same-protocol baseline, so the significance for the community cannot yet be assessed at the level claimed in the abstract.

major comments (4)

[§4, App. A.2] The 5.1 cm straight-path error is the discrepancy between the robot's LiDAR-estimated position and the operator's Apple Vision Pro odometry position, and this same AVP position is the reference signal that the closed-loop policy is trained to chase (Section 3.2, Table A1 'Body position (MR)'). The paper gives no characterization of AVP odometry drift over 3-8.9 m trajectories and no description of the calibration or frame-alignment procedure between the AVP frame and the LiDAR map frame beyond fixing initial positions. If the operator reference drifts, the robot will faithfully track a drifting target, and the reported error measures agreement between two drifting estimators rather than global positioning accuracy. This is load-bearing for the abstract claim of 'minimal positional drift over long-range trajectories'; the authors should report reference drift on the same paths or a stationary AVP test, and describe the frame-alignment procedure.
[§4] The improvement over existing open-loop systems is asserted but not measured. The real-world path-following experiment has no baseline run under the same protocol, e.g., the same policy with the closed-loop error-correction term disabled, or an existing open-loop teleoperation system such as OmniH2O on the same 3/6/8.9 m straight path. Without a same-protocol baseline, the 5.1 cm figure is an absolute performance number; it does not support 'improved tracking precision over existing open-loop approaches' or 'unprecedented' fidelity. Adding an open-loop control condition in the same environment is necessary to substantiate the central contribution.
[Table 1, Fig. 7] The simulation metrics in Table 1 are computed on CLONED, the same dataset used to train the policy, and all methods report 100% success rate, so these numbers quantify in-distribution tracking rather than generalization. Figure 7 uses height-edited sequences from CLONED, which is a mild distribution shift but still within the source motion repertoire. The paper should state the train/test separation explicitly, or evaluate on held-out motion sequences and unseen stance configurations, to support the claim of robust generalization.
[Eq. (2), Table A2] The velocity-dependent SDE noise model is presented as reflecting real LiDAR and IMU error characteristics, but the constants cvel and cmin are not reported in Table A2, and the model adds noise only to the head position, not to the operator's global reference. Slow VIO drift and frame misalignment between operator and robot odometry are precisely the error modes that the closed-loop claim depends on, yet they are absent from the training noise model. Please report the constants and either justify the omission of reference drift in training or include a variant that models it.

minor comments (6)

[Fig. 2] The figure label 'Close-Loop Tacking' contains typos; it should read 'Closed-Loop Tracking'.
[App. A.1] The word 'teloperation' is misspelled; it should be 'teleoperation'.
[§4, Fig. 4] The plot shows mean errors but no error bars or per-condition standard deviations, although each straight-path condition was repeated ten times; reporting spread would help assess consistency.
[App. D.3] The subsection title 'Qualitative Results Comparsion' should be 'Qualitative Results Comparison'.
[§5, Fig. 7] In the legend, 'CLONE †' and 'CLONE *' are rendered with inconsistent spacing; this is cosmetic but should be fixed for clarity.
[§4] The t-tests comparing distance groups have only n=10 per group; the absence of a significant difference is weak evidence of no drift accumulation and should be phrased accordingly.

Circularity Check

1 steps flagged · score 6.0 of 10

Headline drift metric is measured against the same AVP reference the policy is trained to chase; partial circularity.

self definitional [Sec. 3.2 (closed-loop error correction), Sec. 4 (Global Position Tracking), Table A1]
"the teleoperator's position ˆp ∈ R3 is similarly tracked through a MR hardware equipped with a comparable odometry pipeline. The student teleoperation policy directly consumes the difference between p and ˆp, enabling it to generate actions that systematically reduce positional drift; App. A.2: we obtain the operator's global position with the odometry from Apple Vision Pro; Sec. 4: We measured the discrepancy between the robot's final and expected positions as the tracking error; Table A1: Body position (MR) exp(−0.5∥qvr − q̂vr∥²) 150."

The error reported as 5.1 cm mean (12.0 cm max) is the robot-vs-AVP position difference. This same difference is both the policy observation (p − p̂) and the training reward ('Body position (MR)', weight 150). Hence the headline number measures how well the learned controller minimized its own objective; it cannot independently establish 'global positioning' or 'minimal positional drift.' AVP odometry drift and LiDAR/AVP frame alignment are never characterized, so if the AVP reference drifts, the robot chases the drift and the metric stays small. The claim is true by construction relative to the reference the loop is trained to track, not relative to an external ground-truth position.

full rationale

The central quantitative claim (5.1 cm straight-path error, 'minimal positional drift') is partially circular: the metric is exactly the robot-vs-AVP difference that the policy is trained to minimize via its observation and the 'Body position (MR)' reward. Evaluating the controller against its own reference is legitimate for tracking performance but does not validate global positioning unless the AVP reference is independently characterized, which the paper does not do. I do not count Table 1 (evaluation on the CLONED training set) as a circular step, since in-distribution evaluation is weak evidence rather than a construction-level equivalence; the MoE ablations, stance-height tests, and real-world demonstrations provide some independent support. No load-bearing self-citation or imported uniqueness theorem appears: the OmniH2O baseline and FAST-LIO2 odometry are external citations. The partial circularity is confined to the flagship drift metric, so the overall score is 6 rather than higher.

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

The central claim rests on the accuracy of two odometry sources (robot and operator), the realism of the SDE noise model and IsaacGym randomization, and the physical feasibility of AMASS retargeting. The only clearly ad hoc modeling element is the SDE noise model with unreported constants; the other assumptions are standard domain assumptions for sim-to-real robot learning. No new physical entities are introduced; the CLONED dataset and MoE policy are artifacts, not postulated entities. Five hand-chosen free-parameter groups (SDE constants, MoE hyperparameters, history length, reward weights, randomization ranges) are required to reproduce the results.

free parameters (5)

cvel and cmin in SDE noise model = not reported
Equation (2) uses these constants to scale LiDAR odometry noise with movement speed; no fitting procedure or values are given, so they are hand-chosen model inputs.
MoE architecture hyperparameters (L=3 layers, N=4 experts, k=2 top experts) = L=3, N=4, k=2
Selected via ablation (Table A3) rather than derived; the central MoE claim depends on these choices.
History length H = 25 frames
Selected via ablation (Table A3); performance degrades at 5 and 50 frames.
Reward weights in Table A1 = e.g., AMP 500, Body position (MR) 150, hand rotation 500
Hand-tuned training weights; the policy behavior described in the experiments depends on them.
Domain randomization ranges in Table A2 = e.g., friction 0.6-2.0, mass 0.7-1.25x default
Hand-chosen ranges used to make the sim policy transfer to the real robot; not derived from measured distributions.

assumptions (5)

domain assumption FAST-LIO2 LiDAR-inertial odometry yields accurate, drift-bounded global pose estimates for the humanoid in the deployment environment.
Sections 3.2 and A.2 use FAST-LIO2 output as the robot's position feedback; if it drifts or fails on the G1's dynamic motions, the error correction target is wrong.
domain assumption Apple Vision Pro visual-inertial odometry yields an accurate, drift-free operator global position that can serve as the control reference.
Sections 3.2 and A.2; the policy minimizes the difference between robot and operator positions, so drift in the operator reference becomes drift in the robot.
ad hoc to paper The velocity-dependent SDE noise model in Eq. (2) captures real odometry error statistics well enough for sim-to-real transfer.
Equation (2) introduces cvel and cmin without measuring real odometry noise; this is an ad hoc modeling choice for the closed-loop training.
domain assumption IsaacGym simulation with the listed domain randomization adequately models Unitree G1 dynamics for policy transfer.
Appendix C.3 trains in IsaacGym; deployment on the real G1 assumes the sim-to-real gap is covered by randomization.
domain assumption AMASS motions retargeted to the G1 remain physically feasible and cover the lower-body behaviors needed for teleoperation.
Section 3.4 mentions filtering physically infeasible data but gives no quantitative feasibility validation of the retargeted motions.

how reviews work

0 comments

Cite this review

Pith. "Pith review of CLONE: Closed-Loop Whole-Body Humanoid Teleoperation for Long-Horizon Tasks." pith.science (2026). https://pith.science/paper/XJLPL5FD

@misc{pith2026250608931,
  author       = {Pith},
  title        = {Pith review of: CLONE: Closed-Loop Whole-Body Humanoid Teleoperation for Long-Horizon Tasks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XJLPL5FD}},
  note         = {Machine review of arXiv:2506.08931}
}

read the original abstract

Humanoid teleoperation plays a vital role in demonstrating and collecting data for complex humanoid-scene interactions. However, current teleoperation systems face critical limitations: they decouple upper- and lower-body control to maintain stability, restricting natural coordination, and operate open-loop without real-time position feedback, leading to accumulated drift. The fundamental challenge is achieving precise, coordinated whole-body teleoperation over extended durations while maintaining accurate global positioning. Here we show that an MoE-based teleoperation system, CLONE, with closed-loop error correction enables unprecedented whole-body teleoperation fidelity, maintaining minimal positional drift over long-range trajectories using only head and hand tracking from an MR headset. Unlike previous methods that either sacrifice coordination for stability or suffer from unbounded drift, CLONE learns diverse motion skills while preventing tracking error accumulation through real-time feedback, enabling complex coordinated movements such as ``picking up objects from the ground.'' These results establish a new milestone for whole-body humanoid teleoperation for long-horizon humanoid-scene interaction tasks.

Figures

Figures reproduced from arXiv: 2506.08931 by the authors.

**Figure 1.** CLONE employs an MoE-based policy with closed-loop error correction for humanoid teleoperation, enabling precise whole-body coordination and long-horizon task execution. Abstract: Humanoid teleoperation plays a vital role in demonstrating and collecting data for complex humanoid-scene interactions. However, current teleoperation systems face critical limitations: they decouple upper- and lower-body control to maint… view at source ↗

**Figure 2.** Whole-body humanoid teleoperation from minimal input. Our approach enables intuitive control of a humanoid robot using only head and hand poses from mixed reality input, generating coordinated wholebody motions including natural locomotion. Through closed-loop tracking, the system maintains accurate correspondence between operator and robot over extended operation periods, enabling complex long-horizon tasks that… view at source ↗

**Figure 3.** The CLONE framework. (a) CLONED curates and augments retargeted AMASS [28] data through motion editing to introduce diverse humanoid motions and detailed hand movements. (b) A teacher policy is trained using privileged information, including full robot state and environmental context. (c) An MoE network serves as the student policy, distilled from the teacher to operate with real-world observations only. (d) For rea… view at source ↗

Figures from the paper (4 more)

**Figure 4.** Figure 4: Global position tracking accuracy in real-world experiments. CLONE achieves mean tracking errors of 5.1cm across distances up to 8.9m, demonstrating effective closed-loop error correction in extended teleoperation. We evaluated CLONE on a physical Unitree G1 humanoid…

**Figure 5.** Figure 5: Whole-body motion tracking on Unitree G1. CLONE successfully tracks diverse skills including (a) waving, (b)(d) squatting, and (c)jumping, showcasing comprehensive whole-body coordination capabilities. In curved-path tracking, the mean error is 20 cm (maximum 27 cm), a…

**Figure 6.** Figure 6: , we recorded a continuous teleoperation sequence where the operator traversed a complex path spanning over 15m, incorporating diverse locomotion patterns including forward walking, turning, side-stepping, and returning to the original position. The robot consistently …

**Figure 7.** Figure 7: Motion tracking performance across stance heights. Comparison between CLONE (blue solid), CLONE ∗ (green dashed), and CLONE † (red dashed) across different postures from standing to deep squatting. Lower values indicate better performance. Tracking Diverse Stances To a…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 9 Pith papers

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

LapSurgie: Humanoid Robots Performing Surgery via Teleoperated Handheld Laparoscopy
cs.RO 2025-10 conditional novelty 7.0 of 10

A humanoid robot teleoperated via a da Vinci console can grasp off-the-shelf wristed laparoscopic tools and perform a peg-transfer task with accuracy matching the dVRK, but slower.
Flow Matching Policy Gradients
cs.LG 2025-07 conditional novelty 7.0 of 10

FPO trains flow-based policies with PPO by replacing the likelihood ratio with an exponentiated flow matching loss difference.
Visual Imitation Enables Contextual Humanoid Control
cs.RO 2025-05 conditional novelty 7.0 of 10

A single policy trained from 123 monocular videos, fine-tuned in simulation, and distilled to heightmap plus root-direction inputs lets a Unitree G1 climb stairs and sit and stand on real furniture.
HumanTracker: Towards Comprehensive and Human-Aligned Motion Tracking Benchmark
cs.RO 2026-08 conditional novelty 6.0 of 10

HumanTracker introduces a 153-hour categorized humanoid tracking benchmark and a preference-trained metric, HumanScore, that agrees with human judgments better than kinematic error metrics.
HumanoidUMI: Bridging Robot-Free Demonstrations and Humanoid Whole-Body Manipulation
cs.RO 2026-06 unverdicted novelty 6.0 of 10

Robot-free VR–UMI demos of sparse whole-body keypoints can be retargeted and executed as deployable Unitree G1 whole-body skills more efficiently than teleoperation.
Thor: Towards Human-Level Whole-Body Reactions for Intense Contact-Rich Environments
cs.RO 2025-10 conditional novelty 6.0 of 10

A decoupled whole-body RL policy with a force-based lean reward enables a Unitree G1 humanoid to pull with up to 167.7 N, beating prior controllers by 69–75%.
PHUMA: Physically Reliable Humanoid Locomotion Dataset
cs.RO 2025-10 conditional novelty 6.0 of 10

PHUMA is a curated 73-hour humanoid locomotion corpus whose physical-reliability metrics are partly defined by the same losses used to optimize it, and whose imitation success claims are confounded by in-distribution ...
The Latent Color Subspace: Emergent Order in High-Dimensional Chaos
cs.LG 2026-03 unverdicted novelty 5.0 of 10

FLUX.1’s VAE latent space contains an interpretable Hue–Saturation–Lightness structure that enables training-free color prediction and control via closed-form latent edits.
RobotDancing: Residual-Action Reinforcement Learning Enables Robust Long-Horizon Humanoid Motion Tracking
cs.RO 2025-09 conditional novelty 5.0 of 10

Residual-action reinforcement learning, with selective corrections on hip and knee pitch joints, enables zero-shot long-horizon dance tracking on real humanoid robots.

Reference graph

Works this paper leans on

66 extracted references · 66 canonical work pages · cited by 9 Pith papers

[1]

Henze, M

B. Henze, M. A. Roa, and C. Ott. Passivity-based whole-body balancing for torque-controlled humanoid robots in multi-contact scenarios. International Journal of Robotics Research (IJRR), 35(12):1522–1543, 2016

work page 2016
[2]

P. M. Wensing and D. E. Orin. Improved computation of the humanoid centroidal dynamics and application for whole-body control. International Journal of Humanoid Robotics , 13(01): 1550039, 2016

work page 2016
[3]

Sentis and O

L. Sentis and O. Khatib. Synthesis of whole-body behaviors through hierarchical control of behavioral primitives. International Journal of Humanoid Robotics , 2(04):505–518, 2005

work page 2005
[4]

O. Khatib. A unified approach for motion and force control of robot manipulators: The opera- tional space formulation. IEEE Journal on Robotics and Automation , 3(1):43–53, 2003

work page 2003
[5]

Fukuda, P

T. Fukuda, P. Dario, and G.-Z. Yang. Humanoid robotics—history, current state of the art, and challenges. Science Robotics, 2(13):eaar4043, 2017

work page 2017
[6]

Hereid, C

A. Hereid, C. M. Hubicki, E. A. Cousineau, and A. D. Ames. Dynamic humanoid locomotion: A scalable formulation for hzd gait optimization. IEEE Transactions on Robotics (T-RO) , 34 (2):370–387, 2018

work page 2018
[7]

J. Cui, T. Liu, Z. Meng, J. Yu, R. Song, W. Zhang, Y . Zhu, and S. Huang. Grove: A generalized reward for learning open-vocabulary physical skill. InProceedings of Conference on Computer Vision and Pattern Recognition (CVPR), 2025

work page 2025
[8]

Y . Tong, H. Liu, and Z. Zhang. Advancements in humanoid robots: A comprehensive review and future prospects. IEEE/CAA Journal of Automatica Sinica , 11(2):301–328, 2024

work page 2024

Show all 66 references

[9]

Battlefield Extraction-Assist Robot (BEAR)

Vecna Robotics. Battlefield Extraction-Assist Robot (BEAR). https://robotsguide. com/robots/bear, 2006

2006
[10]

Naval Research Laboratory

U.S. Naval Research Laboratory. Autonomous Shipboard Humanoid (ASH). https: //www.navy.mil/Resources/Fact-Files/Display-FactFiles/Article/ 2160601/shipboard-autonomous-firefighting-robot-saffir/ , 2012. 9

2012
[11]

AGIBOT A2 Humanoid Robot

AGIBOT Robotics. AGIBOT A2 Humanoid Robot. https://www.agibot.com/ products/A2, 2024

2024
[12]

Z. Fu, Q. Zhao, Q. Wu, G. Wetzstein, and C. Finn. Humanplus: Humanoid shadowing and imitation from humans. In Conference on Robot Learning (CoRL) , 2024

2024
[13]

Y . Ze, Z. Chen, W. Wang, T. Chen, X. He, Y . Yuan, X. B. Peng, and J. Wu. Generalizable humanoid manipulation with improved 3d diffusion policies. In IEEE/RSJ International Con- ference on Intelligent Robots and Systems (IROS) , 2025

2025
[14]

T. He, W. Xiao, T. Lin, Z. Luo, Z. Xu, Z. Jiang, C. Liu, G. Shi, X. Wang, L. Fan, and Y . Zhu. Hover: Versatile neural whole-body controller for humanoid robots. In IEEE International Conference on Robotics and Automation (ICRA) , 2025

2025
[15]

J. Cui, T. Liu, N. Liu, Y . Yang, Y . Zhu, and S. Huang. Anyskill: Learning open-vocabulary physical skill for interactive agents. In Proceedings of Conference on Computer Vision and Pattern Recognition (CVPR), 2024

2024
[16]

Dafarra, K

S. Dafarra, K. Darvish, R. Grieco, G. Milani, U. Pattacini, L. Rapetti, G. Romualdi, M. Salvi, A. Scalzo, I. Sorrentino, et al. icub3 avatar system: Enabling remote fully immersive embodi- ment of humanoid robots. Science Robotics, 9(86):eadh3834, 2024

2024
[17]

Q. Ben, F. Jia, J. Zeng, J. Dong, D. Lin, and J. Pang. Homie: Humanoid loco-manipulation with isomorphic exoskeleton cockpit. In Robotics: Science and Systems (RSS) , 2025

2025
[18]

T. He, Z. Luo, X. He, W. Xiao, C. Zhang, W. Zhang, K. Kitani, C. Liu, and G. Shi. Om- nih2o: Universal and dexterous human-to-humanoid whole-body teleoperation and learning. In Conference on Robot Learning (CoRL) , 2024

2024
[19]

Haarnoja, B

T. Haarnoja, B. Moran, G. Lever, S. H. Huang, D. Tirumala, J. Humplik, M. Wulfmeier, S. Tun- yasuvunakool, N. Y . Siegel, R. Hafner, et al. Learning agile soccer skills for a bipedal robot with deep reinforcement learning. Science Robotics, 9(89):eadi8022, 2024

2024
[20]

A. Matsiko. Humanoid robot learning of complex behaviors with llms. Science Robotics, 10 (98):eadv4627, 2025

2025
[21]

T. He, Z. Luo, W. Xiao, C. Zhang, K. Kitani, C. Liu, and G. Shi. Learning human-to-humanoid real-time whole-body teleoperation. In IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2024

2024
[22]

Cheng, Y

X. Cheng, Y . Ji, J. Chen, R. Yang, G. Yang, and X. Wang. Expressive whole-body control for humanoid robots. In Robotics: Science and Systems (RSS) , 2024

2024
[23]

M. Ji, X. Peng, F. Liu, J. Li, G. Yang, X. Cheng, and X. Wang. Exbody2: Advanced expressive humanoid whole-body control, 2024

2024
[24]

Jiang, Z

N. Jiang, Z. Zhang, H. Li, X. Ma, Z. Wang, Y . Chen, T. Liu, Y . Zhu, and S. Huang. Scaling up dynamic human-scene interaction modeling. In Proceedings of Conference on Computer Vision and Pattern Recognition (CVPR), 2024

2024
[25]

Jiang, Z

N. Jiang, Z. He, H. Li, Y . Chen, S. Huang, and Y . Zhu. Autonomous character-scene interaction synthesis from text instruction. In SIGGRAPH Asia Conference Papers, 2024

2024
[26]

Jiang, H

N. Jiang, H. Li, Z. Yuan, Z. He, Y . Chen, T. Liu, Y . Zhu, and S. Huang. Dynamic motion blending for versatile motion editing. In Proceedings of Conference on Computer Vision and Pattern Recognition (CVPR), 2025

2025
[27]

W. Xu, Y . Cai, D. He, J. Lin, and F. Zhang. Fast-lio2: Fast direct lidar-inertial odometry.IEEE Transactions on Robotics (T-RO), 38(4):2053–2073, 2022. 10

2022
[28]

Mahmood, N

N. Mahmood, N. Ghorbani, N. F. Troje, G. Pons-Moll, and M. J. Black. AMASS: Archive of motion capture as surface shapes. In Proceedings of International Conference on Computer Vision (ICCV), 2019

2019
[29]

Darvish, Y

K. Darvish, Y . Tirupachuri, G. Romualdi, L. Rapetti, D. Ferigo, F. J. A. Chavez, and D. Pucci. Whole-body geometric retargeting for humanoid robots. In International Conference on Hu- manoid Robots (Humanoids), 2019

2019
[30]

Brygo, I

A. Brygo, I. Sarakoglou, N. Garcia-Hernandez, and N. Tsagarakis. Humanoid robot teleopera- tion with vibrotactile based balancing feedback. InHaptics: Neuroscience, Devices, Modeling, and Applications, 2014

2014
[31]

Peternel and J

L. Peternel and J. Babiˇc. Learning of compliant human–robot interaction using full-body haptic interface. Advanced Robotics, 27(13):1003–1012, 2013

2013
[32]

Ramos and S

J. Ramos and S. Kim. Dynamic locomotion synchronization of bipedal robot and human operator via bilateral feedback teleoperation. Science Robotics, 4(35):eaav4282, 2019

2019
[33]

Chagas Vaz, D

J. Chagas Vaz, D. Wallace, and P. Y . Oh. Humanoid loco-manipulation of pushed carts uti- lizing virtual reality teleoperation. In International Mechanical Engineering Congress and Exposition, 2021

2021
[34]

Penco, N

L. Penco, N. Scianca, V . Modugno, L. Lanari, G. Oriolo, and S. Ivaldi. A multimode teleop- eration framework for humanoid loco-manipulation: An application for the icub robot. IEEE Robotics & Automation Magazine, 26(4):73–82, 2019

2019
[35]

Tachi, Y

S. Tachi, Y . Inoue, and F. Kato. Telesar vi: Telexistence surrogate anthropomorphic robot vi. International Journal of Humanoid Robotics , 17(05):2050019, 2020

2020
[36]

Nakamura, H

Y . Nakamura, H. Hanafusa, and T. Yoshikawa. Task-priority based redundancy control of robot manipulators. International Journal of Robotics Research (IJRR) , 6(2):3–15, 1987

1987
[37]

Khatib, L

O. Khatib, L. Sentis, J. Park, and J. Warren. Whole-body dynamic behavior and control of human-like robots. International Journal of Humanoid Robotics , 1(01):29–43, 2004

2004
[38]

Dietrich, C

A. Dietrich, C. Ott, and A. Albu-Sch¨affer. An overview of null space projections for redundant, torque-controlled robots. International Journal of Robotics Research (IJRR) , 34(11):1385– 1400, 2015

2015
[39]

F. L. Moro and L. Sentis. Whole-body control of humanoid robots. Humanoid robotics: a reference, pages 1161–1183, 2019

2019
[40]

Moniruzzaman, A

M. Moniruzzaman, A. Rassau, D. Chai, and S. M. S. Islam. Teleoperation methods and en- hancement techniques for mobile robots: A comprehensive survey. Robotics and Autonomous Systems, 150:103973, 2022

2022
[41]

Huang, S

R. Huang, S. Zhu, Y . Du, and H. Zhao. Moe-loco: Mixture of experts for multitask locomotion. In IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , 2025

2025
[42]

S. Zhou, W. Zhang, J. Jiang, W. Zhong, J. GU, and W. Zhu. On the convergence of stochastic multi-objective gradient manipulation and beyond. In Proceedings of Advances in Neural Information Processing Systems (NeurIPS), 2022

2022
[43]

Darvish, L

K. Darvish, L. Penco, J. Ramos, R. Cisneros, J. Pratt, E. Yoshida, S. Ivaldi, and D. Pucci. Teleoperation of humanoid robots: A survey. IEEE Transactions on Robotics (T-RO) , 39(3): 1706–1727, 2023

2023
[44]

C. Yang, K. Yuan, Q. Zhu, W. Yu, and Z. Li. Multi-expert learning of adaptive legged locomo- tion. Science Robotics, 5(49):eabb2174, 2020. 11

2020
[45]

Z. Xie, S. Starke, H. Y . Ling, and M. van de Panne. Learning soccer juggling skills with layer-wise mixture-of-experts. In ACM SIGGRAPH / Eurographics Symposium on Computer Animation (SCA), 2022

2022
[46]

W. Song, H. Zhao, P. Ding, C. Cui, S. Lyu, Y . Fan, and D. Wang. Germ: A generalist robotic model with mixture-of-experts for quadruped robot. In IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , 2024

2024
[47]

Cheng, L

G. Cheng, L. Dong, W. Cai, and C. Sun. Multi-task reinforcement learning with attention- based mixture of experts. IEEE Robotics and Automation Letters (RA-L) , 8(6):3812–3819, 2023

2023
[48]

C. R. Garrett, R. Chitnis, R. Holladay, B. Kim, T. Silver, L. P. Kaelbling, and T. Lozano-P´erez. Integrated task and motion planning. Annual Review of Control, Robotics, and Autonomous Systems, 4(1):265–293, 2021

2021
[49]

C. Wang, L. Fan, J. Sun, R. Zhang, L. Fei-Fei, D. Xu, Y . Zhu, and A. Anandkumar. Mimicplay: Long-horizon imitation learning by watching human play. In Conference on Robot Learning (CoRL), 2023

2023
[50]

Z. Lin, Y . Chen, and Z. Liu. Hierarchical human-to-robot imitation learning for long-horizon tasks via cross-domain skill alignment. In IEEE International Conference on Robotics and Automation (ICRA), 2024

2024
[51]

H. Shi, H. Xu, S. Clarke, Y . Li, and J. Wu. Robocook: Long-horizon elasto-plastic object manipulation with diverse tools. Conference on Robot Learning (CoRL) , 2023

2023
[52]

Z. Zhao, Y . Li, W. Li, Z. Qi, L. Ruan, Y . Zhu, and K. Althoefer. Tac-man: Tactile-informed prior-free manipulation of articulated objects. IEEE Transactions on Robotics (T-RO) , 41: 538–557, 2024

2024
[53]

Z. Jiao, Z. Zeyu, W. Wang, D. Han, S.-C. Zhu, Y . Zhu, and H. Liu. Efficient task planning for mobile manipulation: a virtual kinematic chain perspective. In IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , 2021

2021
[54]

Z. Jiao, Z. Zeyu, X. Jiang, D. Han, S.-C. Zhu, Y . Zhu, and H. Liu. Consolidating kinematic models to promote coordinated mobile manipulations. In IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , 2021

2021
[55]

Z. Jiao, Y . Niu, Z. Zhang, S.-C. Zhu, Y . Zhu, and H. Liu. Planning sequential tasks on contact graph. In IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2022

2022
[56]

P. Zhi, Z. Zhang, M. Han, Z. Zhang, Z. Li, Z. Jiao, B. Jia, and S. Huang. Closed-loop open- vocabulary mobile manipulation with gpt-4v. In IEEE International Conference on Robotics and Automation (ICRA), 2024

2024
[57]

P. Zhi, P. Li, J. Yin, B. Jia, and S. Huang. Learning unified force and position control for legged loco-manipulation. In Conference on Robot Learning (CoRL) , 2025

2025
[58]

Y . Su, J. Li, Z. Jiao, M. Wang, C. Chu, H. Li, Y . Zhu, and H. Liu. Sequential manipulation plan- ning for over-actuated unmanned aerial manipulators. In IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , 2023

2023
[59]

Wisth, M

D. Wisth, M. Camurri, and M. Fallon. Vilens: Visual, inertial, lidar, and leg odometry for all-terrain legged robots. IEEE Transactions on Robotics (T-RO), 39(1):309–326, 2022

2022
[60]

G. Ou, D. Li, and H. Li. Leg-kilo: Robust kinematic-inertial-lidar odometry for dynamic legged robots. IEEE Robotics and Automation Letters (RA-L) , 2024. 12

2024
[61]

Allshire, H

A. Allshire, H. Choi, J. Zhang, D. McAllister, A. Zhang, C. M. Kim, T. Darrell, P. Abbeel, J. Malik, and A. Kanazawa. Visual imitation enables contextual humanoid control. In Confer- ence on Robot Learning (CoRL) , 2025

2025
[62]

F. G. Harvey, M. Yurick, D. Nowrouzezahrai, and C. Pal. Robust motion in-betweening. ACM Transactions on Graphics (TOG), 39(4):60–1, 2020

2020
[63]

L. Ma, Z. Meng, T. Liu, Y . Li, R. Song, W. Zhang, and S. Huang. Styleloco: Generative adversarial distillation for natural humanoid robot locomotion, 2025

2025
[64]

H. Geng, F. Wang, S. Wei, Y . Li, B. Wang, B. An, C. T. Cheng, H. Lou, P. Li, Y .-J. Wang, et al. Roboverse: Towards a unified platform, dataset and benchmark for scalable and generalizable robot learning. In Robotics: Science and Systems (RSS) , 2025

2025
[65]

Z. Wang, Y . Chen, T. Liu, Y . Zhu, W. Liang, and S. Huang. Humanise: Language-conditioned human motion generation in 3d scenes. In Proceedings of Advances in Neural Information Processing Systems (NeurIPS), 2022

2022
[66]

crouch” or “squat to pick up an object

X. B. Peng, Z. Ma, P. Abbeel, S. Levine, and A. Kanazawa. Amp: Adversarial motion priors for stylized physics-based character control. ACM Transactions on Graphics (TOG) , 40(4): 1–20, 2021. 13 A Preliminaries A.1 Formulation We formulate the humanoid teloperation as a Markov ...

2021

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [1]

Henze, M

B. Henze, M. A. Roa, and C. Ott. Passivity-based whole-body balancing for torque-controlled humanoid robots in multi-contact scenarios. International Journal of Robotics Research (IJRR), 35(12):1522–1543, 2016

work page 2016

[2] [2]

P. M. Wensing and D. E. Orin. Improved computation of the humanoid centroidal dynamics and application for whole-body control. International Journal of Humanoid Robotics , 13(01): 1550039, 2016

work page 2016

[3] [3]

Sentis and O

L. Sentis and O. Khatib. Synthesis of whole-body behaviors through hierarchical control of behavioral primitives. International Journal of Humanoid Robotics , 2(04):505–518, 2005

work page 2005

[4] [4]

O. Khatib. A unified approach for motion and force control of robot manipulators: The opera- tional space formulation. IEEE Journal on Robotics and Automation , 3(1):43–53, 2003

work page 2003

[5] [5]

Fukuda, P

T. Fukuda, P. Dario, and G.-Z. Yang. Humanoid robotics—history, current state of the art, and challenges. Science Robotics, 2(13):eaar4043, 2017

work page 2017

[6] [6]

Hereid, C

A. Hereid, C. M. Hubicki, E. A. Cousineau, and A. D. Ames. Dynamic humanoid locomotion: A scalable formulation for hzd gait optimization. IEEE Transactions on Robotics (T-RO) , 34 (2):370–387, 2018

work page 2018

[7] [7]

J. Cui, T. Liu, Z. Meng, J. Yu, R. Song, W. Zhang, Y . Zhu, and S. Huang. Grove: A generalized reward for learning open-vocabulary physical skill. InProceedings of Conference on Computer Vision and Pattern Recognition (CVPR), 2025

work page 2025

[8] [8]

Y . Tong, H. Liu, and Z. Zhang. Advancements in humanoid robots: A comprehensive review and future prospects. IEEE/CAA Journal of Automatica Sinica , 11(2):301–328, 2024

work page 2024