Dexterous and contact-rich robotic manipulation strategies.
Jiaxin Shi, Xidong Zhang, Fucai Zhu et al.
We propose a 3D-thinking-guided co-training framework that enables vision-language-action (VLA) models to perform 3D spatial reasoning implicitly during action prediction.
Octi Zhang, Mateo Guaman Castro, Patrick Yin et al.
General-purpose robots must perform a wide range of tasks from agile locomotion to dexterous manipulation.
Mozes Jacobs, Aiden Swann, Mathilde Papillon et al.
Robotic manipulation is inherently geometric, but what shape does the world take inside the policy that performs it?
Chen Cai, Steven Liu
Close-proximity multi-arm manipulation requires collision models that are both geometrically accurate and differentiable enough for real-time optimization.
Mathilde Kappel, Clémence Grislain, Mohamed Chetouani et al.
Vision-Language-Action (VLA) models have become a prominent paradigm for mapping multimodal inputs, including semantic instructions, visual observations of the scene, and proprioceptive observations, to robot actions.
Dang Viet Anh Nguyen, Rajesh Piplani, Aldy Gunawan
We introduce the two-echelon covering tour vehicle routing problem (2E-CTVRP) for the distribution of relief supplies after a disaster. In the first echelon, a fleet of trucks transports supplies and drones from a central depot to satellites at the periphery of the affected area.
Junheng Li, Liang Wu, Sergio A. Esteban et al.
In humanoid motion control, model predictive control (MPC) offers physically grounded prediction and constraint handling, while reinforcement learning (RL) enables robust whole-body skills through large-scale simulation. However, using MPC inside RL often requires time-consuming problem construction or excessive training overhead, making such frameworks difficult to justify in practice. This work studies efficient training-time MPC guidance for humanoid locomotion and manipulation, termed MPC-RL. We introduce a centroidal-dynamics MPC reward formulation that leverages guidance from MPC trajectories in training time. To make this practical in massively parallel RL, we develop π^nMPC, a parallel-in-horizon and construction-free batched GPU MPC solver that operates directly on time-varying dynamics to avoid high memory usage and pre-compilation. Through a variety of comparative studies and hardware validations, we have found that MPC-RL achieves superior performance in locomotion and manipulation skills.
Jiabin Qiu, Zixuan Chen, Hongye Cao et al.
Latent world models are typically trained to predict factual transitions, whereas model predictive control (MPC) must compare alternative actions from the same state. A model can therefore achieve low factual prediction error yet poorly distinguish candidate actions.
Sixu Yan, Shikang Wang, Binhua Huang et al.
This paper proposes AdaRoboVLG, a task-adaptive Vision-Language-Grasp (VLG) framework that supports generalizable grasp synthesis across different robotic hands.
Seok Joon Kim, Junho Lee, Federica Spinola et al.
Direct hand-driven teleoperation maps an operator's hand motion to robot end-effector commands at every frame, enabling precise control, but it requires constant monitoring and correction during approach, grasp, and placement, which can be slow and fatiguing.
Jianli Sun, Bin Tian, Qiyao Zhang et al.
Aerial manipulation systems have long suffered from representation coupling in end-to-end control, as platform-level Unmanned Aerial Vehicle (UAV) movement and end-effector-level arm manipulation differ substantially in action scale, dynamics, and control objectives.
Teeratham Vitchutripop, Alyssa Quarles, Wenhe Zhang et al.
Over the course of a lifetime, robots may encounter novel scenarios unaccounted for in its original training that result in performance degradation. One common approach to mitigating this issue is to further grow the offline training dataset in hopes of producing a policy robust to these changes.
Andrew S. Morgan, Gregory Xie, Capprin Bass et al.
AthenaZero is a bimanual manipulator designed to minimize inertia without compromising control authority.
Rajneesh Anand, Mayuresh V. Kothare
Self-driving laboratories (SDLs) are transforming chemical and materials discovery through closed-loop automation, yet automated infrastructure for physical manipulation of soft, deformable matter remains beyond current robotic platforms.
Mengfei Zhao, Dihong Huang, Yikai Tang et al.
Learning effective robot manipulation policies requires diverse, high-quality demonstrations, yet existing data pipelines are often difficult to scale because they rely on specialized hardware, centralized operators, or fixed task suites.
Seohong Park, Sergey Levine
Behavioral cloning (BC), despite its simplicity, exhibits many counterintuitive phenomena in the real world. For example, the performance of BC often keeps increasing as the model overfits more to the dataset, and fully closed-loop policies often completely fail without action chunking.
Jiahe Pan, Stelian Coros, Jitendra Malik et al.
A primary bottleneck in contact-rich manipulation is the difficulty of collecting real-world data. Sim-to-real reinforcement learning offers a scalable alternative, but the simulation-reality gap prev
Haoran Chen, Jingtian Ji, Samuel Wheeler et al.
Autoregressive action-token policies such as vision-language-action models require action tokenizers to translate discrete token sequences into precise control actions in continuous space. Many action tokenizers learn the mapping between tokens and actions via a reconstruction objective.
Namiko Saito, Hiroshi Kera
Manipulation requires not only reasoning about the external environment, but also about the robot's physical condition.
Zhongxi Chen, Yifan Han, Yanming Shao et al.
Vision-Language-Action (VLA) models have emerged as a promising paradigm for grounding visual-language understanding into real-world robotic manipulation. However, dexterous manipulation remains chall
Xue Qin, Simin Luan, Cong Yang et al.
Sim-to-real research pursues physics fidelity as a primary objective: simulators are judged by how closely they reproduce real-world contact dynamics.
Zongyuan Zhang, Christopher Lehnert, Will N. Browne et al.
Forceful manipulation is challenging for humanoid robots because interaction forces can disturb whole-body balance. We introduce the Supporting Hand Strategy (SHS), which enables a humanoid to brace against the environment with one hand while performing forceful manipulation with the other.
Philipp Ausserlechner, Bernhard Neuberger, Alessandro Scherl et al.
Detecting and grasping small objects remains a significant challenge in robotics. Active vision, where the robot moves closer to the object, is an intuitive solution, yet comparing approaches on common ground is difficult since identical physical scene setups are required.
Hu, Jun, Chen, Sihan, Jovanovic, Kosta et al.
Most robots manipulate objects solely with their end effectors, whereas humans flexibly leverage different body parts, such as the forearm and elbow, especially when handling oversized objects.
Zhaoda Du, Michael Bowman, Xiaoli Zhang
Shared control often allocates robot authority based on confidence in inferred human intent, assuming reliable autonomous execution. When this assumption fails, high intent confidence can cause over-helping.
Yinlong Dai, Homanga Bharadhwaj, Dylan P. Losey
Prior research on learning from humans often requires real human demonstrations. But advances in video generation models offer a new paradigm: we can now prompt these generators to make synthetic videos of the task.
Aiden Swann, Mathilde Papillon, Mozes Jacobs et al.
Vision-language-action (VLA) models are a promising approach to general purpose robot manipulation.
Shutong Ding, Zhaoyang Liu, Yimiao Zhou et al.
Diffusion policies are increasingly used in robotic control and manipulation, vision-language-action models, and world-model-based agents. Their ability to represent multimodal action distributions offers rich opportunities for exploration in reinforcement learning (RL).
Tianjiao Yu, Zhilin Ou, Yifan Shen et al.
Generative simulation can automate robot environment construction, but physically valid environments are not necessarily useful for training. The challenge is to identify environments that address a policy's failures and supervision that helps it overcome them.
Shaocheng Liu, Yuan Meng, Zhenshan Bing et al.
Learning a collective policy for state-based Cartesian manipulation across diverse robotic embodiments presents a fundamental challenge due to embodiment-dependent transition dynamics.
Al Jaber Mahmud, Shuai Li, Xuan Wang
In this paper, we study the joint selection of an environmental support contact and a whole-body configuration for a prescribed loco-manipulation task. A contact may provide greater physical support while restricting the motion required for the task.
Jiayun Li, Georgia Chalvatzaki
Contact-rich motion planning (CRMP) is essential for robotic manipulation and locomotion, yet remains computationally challenging due to combinatorial contact decisions.
Gehan Zheng, Matthew Johnson-Roberson, Weiming Zhi
Contact-rich manipulation failures are often detected only after the robot has committed to contact.
Xinyao Li, Xialin He, Runpei Dong et al.
Keypoint tracking alone is insufficient for object interaction tasks such as sitting on a chair, wiping a board, or pushing furniture, where the robot can reach the correct pose without making meaningful physical contact with the object.
Zhiwei Xue, Jia Yue Kam, Jinhang Qiu et al.
Diffusion models can represent complex, multimodal trajectory distributions, but extracting uncertainty from them typically requires costly Monte Carlo sampling. This limits their use in real-time control, where robots must rapidly assess risk and maintain safety margins.
Hexian Ni, Tao Lu, Yinghao Cai
CoRe combines code-based rewards with learned visual preferences so that neither reward source has to capture the entire task alone.
Mehmet Turan Yardımcı
Multi-objective reinforcement learning for humanoid robots must coordinate locomotion and manipulation within a single policy.
Huang Huang, Sriram Yenamandra, Arjun Majumdar et al.
The diversity of robot embodiments and action spaces makes it challenging to build robot world models that generalize across different embodiments.
Atiksh Bhardwaj, Edward Weiyi Duan, Prithwish Dan et al.
Vision-language-action models (VLAs) are moving rapidly towards deployment as general-purpose manipulation policies, but we currently lack basic tools for understanding what these models represent internally or for monitoring them at runtime.
Siyuan Ma, Boshi Zhang, Yutian Zhang et al.
Mobile manipulation requires a robot to predict how locomotion and arm motion jointly alter future observations and control. Existing world-action models, developed largely for fixed-base platforms, do not explicitly distinguish camera ego-motion from base and arm actions.
Kosar Behnia, Ville Kyrki, Gokhan Alcan
Manipulating deformable objects (DOs) is challenging due to their high-dimensional state space, underactuated dynamics, and partial observability.
Shivam Vats, Sudarshan Harithas, Mete Tuluhan Akbulut et al.
We consider the problem of autonomously learning robot skills under a limited practice budget for sequential tasks.
Juan José García Cárdenas, Alperen Kenan, Hamidreza Raei et al.
Intuitive teleoperation interfaces are crucial for the safe and effective operation of robotic manipulators in challenging environments.
Thang Tran Viet, Thanh Nguyen Canh, Huy Uong Gia et al.
Human-robot interaction (HRI) enables intuitive and intelligent collaboration between humans and robots in real-world environments. This paper introduces a humanoid robot prototype designed as a flexible testbed for developing and integrating artificial intelligence (AI) modules in HRI tasks.
Liang Zhou, Jiaming Su, Yancong Wei et al.
Quadruped manipulators enable mobile grasping in complex environments, yet their whole-body control policies remain vulnerable to unreliable onboard visual perception.
Jusuk Lee, Sungha Kim, Yeonsoo Park et al.
While learning dexterous manipulation from a single human video offers a promising alternative to costly robot demonstrations, many recent methods predominantly imitate demonstrated motions.
Amber Xie, Jiawei Fu, Xiaofeng Guo et al.
Collecting high-quality demonstrations remains a central bottleneck for learning dexterous manipulation. Teleoperation records executable robot actions, but it requires a complete robot setup and places the burden of coordinating many finger joints on the operator.
Botao He, Yiming Dou, Yiannis Aloimonos et al.
How can robots autonomously collect the contact-rich experience needed to learn dexterous manipulation?
Yunchao Yao, Zhuxiu Xu, Tianqi Zhang et al.
Building general-purpose dexterous manipulation policies requires benchmarks that go beyond isolated tasks to systematically evaluate policies across diverse interaction modes, sensory conditions, and robot embodiments.
Pranav Wagh, Yu Fang, Yue Yang et al.
Long-horizon robotic manipulation is often built by chaining independently trained skills.
Joaquin Palacios, Katelyn Lee, Cheng Zhang et al.
Collecting data for manipulation with high-DOF hands is challenging, as interfaces must capture rich hand motion while rendering the contact interactions essential for precise manipulation.
DreamX Team, Rui Chen, Xiangxiang Chu et al.
We present \textbf{DreamX-Phi 1.0}, an action-conditioned video world model for robotic manipulation that, given an observed frame, a language instruction, and a prescribed action sequence comprising end-effector poses and gripper states, predicts the resulting future observations.
Youngjun Jun, Kyumin Choi, Youngmin Kim et al.
Vision-Language-Action (VLA) models increasingly rely on action experts that generate short action chunks under receding-horizon control.
Alexey Zemtsov, Maxim Bobrin, Alexander Nikulin et al.
Offline goal-conditioned reinforcement learning requires both long-horizon reachability estimates and local action comparisons. Dual goal representations provide value fields that capture global goal reachability, but they do not directly specify which action should be preferred at a given state.
Kapalik Khanal, Biswash Khatiwada, Stephen Afrifa et al.
Changing a humanoid robot's manipulation policy should not require retraining the controller that maintains the robot's balance.
Hanchu Zhou, Dechen Gao, Hang Wang et al.
Vision-language models (VLMs) and vision-language-action models (VLAs) have recently driven rapid progress in general-purpose robots, yet most progress has focused on single-robot settings.
Jusuk Lee, Seungjae Lee, Jonghun Shin et al.
Robot manipulation critically depends on perception that preserves the action-relevant aspects of a scene. Yet most robot learning pipelines are built upon visual encoders pre-trained for static recog
Junfeng Li, Junjie He, Zhide Zhong et al.
Vision-Language-Action (VLA) models have become a powerful paradigm for robot manipulation, but training a single generalist policy for heterogeneous robot embodiments remains an open problem. Existing methods have two main limitations.
Jiajian Zheng, Bin Zhang, Wen Fan et al.
Multi-finger hands contact objects at different times, so a global close command can overload early-contact fingers while others are still moving. Dense tactile arrays address this problem but add hardware, calibration, bandwidth, and computation.
Duo Zhang, Zhizhuo Zhang, Xiaoli Bai et al.
We present FAVOR (Free-floating Arm Velocity Optimization with Recursive Sensitivities), a planner for collision-free reaching, tracking, and prescribed-time pre-grasp interception on an unactuated spacecraft.
Christen Millerdurai, Shaoxiang Wang, Yaxu Xie et al.
Reconstructing the absolute 3D pose and shape of the hands from the user's viewpoint using a single head-mounted camera is crucial for practical egocentric interaction in AR/VR, telepresence, and hand-centric manipulation tasks, where sensing must remain compact and unobtrusive.
Tsunehiko Tanaka, Matthew Stephenson, Alistair Macvicar et al.
Robots executing long-horizon manipulation tasks from natural-language instructions must reason about both semantic task structure and geometric feasibility. However, under partial observability, the availability of goal-relevant objects may be uncertain.
Lifeng Zhuo, Wendi Chen, Han Xue et al.
In contact-rich manipulation, action multimodality and reactivity dominate different stages of a single episode. Before contact, multiple trajectories might be equally valid, making it important to preserve diverse action modes.
Shiyuan Yang, Borong Zhang, Jizheng Zhang et al.
We present FabriVLA, a lightweight Vision-Language-Action model for Precise Multi-Task Manipulation.
Zhi Cao, Yining Hong, Jiaolong Yang et al.
Human demonstrations offer a scalable alternative to costly robot data collection for dexterous manipulation, but most human datasets lack force measurements.
Zhengming Zhu, Yang Liu, Xiao Gao et al.
Robot throwing has emerged as a promising technique for improving efficiency in logistics and warehouse automation, by enlarging the workspace and speeding up the process. To significantly increase the throwing system's throughput, we develop strategies for throwing multiple objects in one swipe.
Xiaofan Lu, Kaiji Huang, Jiahui Chen et al.
Curved tactile fingertips for dexterous manipulation must resolve fine contact geometry, distinguish normal and tangential loads, and capture transient signals.
S. Talha Bukhari, Austin Garrett, Yi Wei et al.
Visuomotor policies learn a direct map from raw sensory observations to robot action sequences. Policies based on Diffusion and Flow Matching capture the multimodal distribution over action sequences in an end-to-end manner.
Pavel Bushuyeu, Yujin Chen, Anton Nikolaev et al.
Existing benchmarks evaluate tabletop manipulation, flat-floor household activity, or humanoid locomotion and manipulation as separate task groups; none scores vertical mobility and dexterous work on a fragile payload in one long-horizon episode.
Gia-Binh Nguyen, Trong-Bao Ho, Thien-Loc Ha et al.
Vision-Language-Action (VLA) models pre-trained on massive video-robot datasets have revolutionized robotic manipulation, yet their multi-billion parameter architectures impose prohibitive computational burdens during downstream fine-tuning and real-time inference.
Gia-Binh Nguyen, Trong-Bao Ho, Thien-Loc Ha et al.
Vision-Language-Action (VLA) models pre-trained on massive video-robot datasets have revolutionized robotic manipulation, yet their multi-billion parameter architectures impose prohibitive computational burdens during downstream fine-tuning and real-time inference.
Arman Akbari, Ci Zhang, Arash Akbari et al.
World-action models (WAMs) jointly generate future video and robot actions through iterative diffusion, achieving strong performance on manipulation benchmarks but requiring tens of denoising steps, a cost that precludes real-time control.
Zekai Li, Jiaming Tang, Zhijian Liu
Vision-Language-Action (VLA) models are increasingly promising for robotic manipulation, yet their real-world deployment remains bottlenecked by high inference latency and unstable asynchronous execution.
Jaewoo Lee, Jeongyeon Seo, Sihyun Cho et al.
Embodied coding agents construct robot policies from natural-language instructions by generating executable programs and refining them from execution feedback.
Christos Ziakas, Alessandra Russo, Avishek Joey Bose
Generative policies based on expressive model classes, such as diffusion-models and flow matching, are well-suited to complex control problems with highly multimodal action distributions.
Chenhuan Liu, Yi Xu, Feng Wu et al.
Embodied AI, including vision-language-action and world-action models, must operate reliably in the physical world.
Xiao Tan, Yixuan Wang, Zhexi Wang et al.
FORCA lets a coding agent use wrist force/torque readings both while its robot program runs and when it revises that program between real-robot attempts, without demonstrations or task-specific training.
Zhengyu Tao, Xin Li, Xin Wang
Vision-language-action (VLA) models have demonstrated strong capabilities in robotic manipulation, yet their reliance on visual perception limits robustness in contact-rich environments, where critical physical interaction states may not be visually observable.
Sichang Su, Benjamin Yang, Zhiyun Deng et al.
A pretrained robot foundation policy may execute most of a long-horizon task yet repeatedly fail at a few critical subtasks. Collecting additional full-task demonstrations for supervised fine-tuning (SFT) requires operators to repeat behaviors the policy already performs well.
Ruitong Tian, Fang Xu, Noah B. Wilson et al.
Unlike conventional teleoperation, wearable interfaces allow operators to collect dexterous demonstrations through their own hand motions while directly interacting with task objects.
Dechen Gao, Yue Yang, Ben Abbatematteo et al.
Human demonstrations are a scalable data source for learning dexterous manipulation, but the embodiment gap prevents human motion from being executed directly on robots. Inverse kinematics (IK) retargets human motion to robots efficiently but ignores dynamics, often producing infeasible motions.
Zhihai Bi, Qiang Zhang, Guoyang Zhao et al.
Humanoid-Object Interaction (HOI) is a fundamental capability for humanoid robots, yet it remains challenging due to the tight coupling between dynamic balance and stable interaction with diverse objects.
Chenghao Gu, Hanyang Yu, Jingbo Zhang et al.
Generalist robot policies exhibit strong capabilities, but their robustness in complex and unseen environments remains limited. Scaling robot learning and evaluation in diverse real-world environments remains costly and challenging.
Yifan Xie, Hekun Tian, Jinkun Liu et al.
Human videos provide rich manipulation experience, but extracting action representations that preserve useful motion remains challenging. Visual reconstruction alone can entangle manipulation-related motion with appearance changes and camera movement.
Wenxuan Guo, Ziyuan Li, Meng Zhang et al.
GesVLA augments standard VLA models with gesture awareness, enabling robots to interpret verbal instructions alongside human hand gestures for disambiguated manipulation.
Yupeng Zheng, Xiang Li, Songen Gu et al.
Vision-language pre-training and predictive world modeling provide robot policies with rich semantic and dynamic visual features, but their native action and visual-prediction objectives may omit critical physical and task structure while retaining control-irrelevant visual redundancy.
Yulin Liu, Lai Wei, Yen-Jen Wang et al.
Foundation models and large-scale human data provide rich sources of manipulation intent, but translating this intent into multi-fingered robot behavior remains difficult. Dexterous hands still lack a reusable low-level primitive that reliably establishes contact across tasks and embodiments.
Tianyi Xie, Haotian Zhang, Jinhyung Park et al.
Scaling humanoid loco-manipulation requires robot-compatible demonstrations across diverse objects, whole-body motions, and scene geometries, but teleoperation and motion capture are difficult to scale because each collection depends on physical setups, instrumented actors, and robot operation.
Antony Thomas, Fulvio Mastrogiovanni, Marco Baglietto
Humanoid and service robots operating in clut- tered human environments may encounter motion planning problems that are infeasible when all surrounding objects are treated as fixed obstacles.
Wun Lam Yeung, Wenjun Liu, Yui Cheung Yu et al.
Bimanual object reorientation - picking an object, handing it over between two arms, and placing it in a desired target pose - is valuable when direct placement from the initial grasp is infeasible due to collisions, kinematic constraints, or poor final orientation.
Julien Vanderheyden, Guillaume Drion, Fulvio Forni et al.
Dexterous robot hands must often grasp objects whose shape, size, and pose are known only approximately. Grasp planners typically require accurate object models or correct errors with feedback, but how much inaccuracy a closing motion can tolerate on its own remains unclear.
Hongxin Zhang, Chunru Lin, Junyan Li et al.
Creating dynamic and physically realistic 4D worlds from natural language descriptions is both fascinating and challenging. Traditional computer graphics methods rely on manual creation, requiring extensive human effort to fine-tune materials, motions, and visual fidelity.
Xiaosong Jia, Bowen Yang, Zuhao Ge et al.
Vision-Language-Action (VLA) models aim for general robot learning by aligning action as a modality within powerful Vision-Language Models (VLMs). Existing VLAs rely on end-to-end supervision to implicitly enable the action decoding process to learn task-relevant features.
Jinu Pahk, Jesoon Kang, Taegeon Park et al.
Vision-language-action policies benefit from geometric supervision, but current-frame geometry alone does not explicitly describe the changes associated with manipulation.
Qiuxuan Feng, Jiale Yu, Jiaming Liu et al.
World Action Models (WAMs) have emerged as a promising paradigm for robot control by modeling physical dynamics. Current WAMs generally follow two paradigms: generalizable but imprecise models learned from diverse data, or precise but specialized models fitted to specific tasks.
SenseNova, Joint Research Team, : et al.
GPT-6-Astra demonstrates a level of competence in robotic manipulation and navigation that challenges the conventional paradigm of building specialized models for individual tasks and embodiments.
Zengjue Chen, Peidong Liu, Jiawei Li et al.
Generalist robot policies have demonstrated strong generalization across robotic manipulation tasks, yet their success rates remain limited in com- plex long-horizon scenarios.
Toan Nguyen, Weiduo Yuan, Siheng Zhao et al.
Object transportation is a fundamental capability for humanoid robots operating in real-world, human-centric environments, yet existing methods struggle when clutter constrains free space around both the robot and its carried payload.
An Dang, Arturo Flores Alvarez, Yu-Ming Chen et al.
Humanoid loco-manipulation of large, heavy objects demands forceful interaction across the entire body. However, such payloads shift a humanoid's center of mass and impose sustained loads across the upper body, challenging balance and command tracking.
Ryohei Kobayashi, Kosei Isomoto, Yuga Yano et al.
We propose a framework that transfers tacit human subjective criteria to robotic systems for the appropriate grasping of deformable objects. Achieving such behavior is challenging because a semantic gap exists between qualitative human expectations and quantitative robotic measurements.
Chi Zhang, Penglin Cai, Ziheng Xi et al.
TTP transfers tactile experience from human demonstrations into robot policies for contact-sensitive manipulation.
Yangzhi Yang, Xiansheng Lin, Zhaoming Xie et al.
Humanoid robots could transport payloads substantially heavier than themselves by pulling passive wheeled vehicles instead of carrying the load.
Kyochul Jang, Seohyeon Park, Ohchul Kwon et al.
As robotic hardware and learning methods advance, humanoids need tools to perform tasks beyond their inherent physical limits. Successful tool use requires selecting a suitable tool and coordinating manipulation and, when needed, locomotion to complete the task.
Vincenzo Guarino, Emanuele Musumeci, Vincenzo Suriani et al.
Agentic AI based on Large Language Model generalization capabilities offers a wide range of potential applications, including planning for embodied tasks. For example, embodied agents based on Foundation models can generate plausible plans in autonomous robotics scenarios.
Thomas Landeg, Thierry Tambe
Vision-Language-Action (VLA) models perform well on robotic manipulation, yet remain brittle once their own actions carry them into states their training data barely covers.
Andy Tang, William Chen, Andrew Wagenmaker et al.
Generalist policies can learn a wide range of skills from diverse robot datasets. In order to solve or improve on challenging news tasks, we need a way to infer and invoke the appropriate actions from the policy's rich behavioral prior, especially when directly commanding the policy fails.
Haojian Huang, Zexi Li, Junhao Guo et al.
General-purpose robots must infer what a new task requires and translate that understanding into appropriate physical action.
Yiqi Wang, Zhifeng Rao, Jiaqi Zhang et al.
Open robot policies increasingly follow two paradigms: vision-language-action models (VLAs) directly map observations and instructions to actions, whereas world-action models (WAMs) incorporate learned video or world dynamics into policy learning or action generation.
Zhuo Lin, Sirui Xu, Liuyu Bian et al.
We study test-time evolution for humanoid loco-manipulation: solving tasks that a controller was never trained for by repurposing its existing skills, improving from its own attempts, and retaining what it learns, without retraining.
Chuyang Xiao, Peilin Meng, David Held
Coordinated bimanual manipulation is challenging because the motion of either arm can alter the shared 3D scene and thereby affect the other arm.
Junyi Hu, Zhewen He, Zhenhua Li et al.
A robot demonstration records two things in the same frames: what happened to the objects, and how one particular arm made it happen. We condition on the first.
Yuming Fu, Weijia Wu, Jing Chen et al.
Self-driving laboratories (SDLs) aim to automate the full experimental loop, from scientific decision-making to physical execution.
Siheng Zhao, Yuanhang Zhang, Ziqi Lu et al.
Humanoid robots hold great promise for operating in human-centered environments, yet ladder climbing remains one of the most challenging tasks due to sparse footholds and handholds, complex whole-body coordination, and sensitivity to perception and control errors.
Han Fang, Yunpeng Jiang, Jianshu Hu et al.
Privileged 3D supervision uses additional geometric information during training to guide RGB-based visuomotor policy learning, without requiring geometric inputs at deployment. However, low reconstruction error does not ensure that visual representations capture geometry useful for control.
Zhu, Xinghao, Liu, Zixi, Jain, Shalin et al.
Dexterous robot manipulation can benefit from the abundance of human demonstrations, but transferring such demonstrations to robot policies remains challenging.
Zhixu Li, Keqian Tang, Litian Gong et al.
Personalization is essential for general-purpose robots to transition into household environments and achieve true practicality.
Zuxing Lu, Ziang Zheng, Yao Lyu et al.
Building a general-purpose whole-body controller is essential for enabling diverse motion capabilities in humanoid robots across a wide range of downstream tasks, including locomotion and loco-manipulation.
Yun Zhang, Zhiyu Huang, Tianhui Cai et al.
Long-horizon robotic manipulation fundamentally requires memory: agents must retain task-relevant information from past observations and interactions and use it to guide future actions.
Ke Zhang, Danica J. Sutherland, Chao Liu
Training data variation, whether through designing a domain randomization (DR) scheme in simulation or curating demonstrations for imitation learning, is a primary lever for improving the robustness of robotic manipulation policies.
Sizhe Zhao, Haozhe Xie, Weiyu Zhao et al.
Mainstream robotic policies often adopt a Markovian formulation, but many complex real-world manipulation tasks are inherently non-Markovian, requiring long-horizon memory beyond the current observation.
Sizhe Yang, Juncheng Mu, Tianming Wei et al.
Robust robotic manipulation in the real world requires not only an understanding of the current observation, but also memory and dynamics modeling.
Sizhe Yang, Juncheng Mu, Tianming Wei et al.
Robust robotic manipulation in the real world requires not only an understanding of the current observation, but also memory and dynamics modeling.
Video
Jiye Lee, Yonghun Choi, Jungdam Won
Two-stage framework that enhances noisy multi-human object interaction (MHOI) data by generating plausible hand grasps and refining full-body motion via diffusion-based optimization.
Yajie Li, Bozhou Zhang, Chun Gu et al.
Video foundation-models models offer a promising imagination mechanism for robot manipulation by predicting long-horizon future observations, but effectively exploiting these imagined futures for action execution remains challenging.
Sandeep Bajamahal, Lawrence Yunliang Chen, Toru Lin et al.
Bimanual coordination is essential for many real-world manipulation tasks, yet learning bimanual robot policies is limited by the scarcity of bimanual robots and datasets. Single-arm robots, however,
Tara Sadjadpour, Siming He, C. K. Wolfe et al.
Human hand-object interactions (HOIs) provide a rich source of demonstrations for dexterous manipulation, but learning directly from them presents challenges in bridging morphology gaps, ensuring dynamical feasibility, and sim-to-real deployment.
Jiaxu Wang, Junhao He, Jingkai Sun et al.
MoSA closes the real-to-sim gap for continuum robots by learning a residual stress field that captures anisotropic material behavior missed by standard simulators.
Zirun Ma, Jindou Jia, Tuo An et al.
In robotic manipulation, the same observation often admits multiple feasible actions. Generative policies such as diffusion and flow matching naturally capture such multimodal behaviors, yet which mode is executed is left to random sampling and cannot be actively controlled.
Wei Wang (Robostreet research team. Cambridge, MA, USA) et al.
Refuse collection in pedestrian campuses, parks and mixed-use districts remains manual because compaction trucks cannot reach bins located on walkways, courtyards and building service doors.
Qihang Zhang, Lin Li, Luyao Zhang et al.
The advent of video-action models offers a promising path for robot control. Nevertheless, we argue that repurposing video generative models designed for digital content creation is inherently inadequate for physical environments.
Kinam Kim, Namiko Saito, Heecheol Kim et al.
Vision-Language-Action (VLA) models can generalize across diverse manipulation tasks, but their imitation-learning-based policies remain brittle in precise physical interactions due to compounding execution errors; Can a reinforcement learning policy trained purely in simulation improve the robustness of real-world VLAs zero-shot?
Kinam Kim, Namiko Saito, Heecheol Kim et al.
Vision-Language-Action (VLA) models can generalize across diverse manipulation tasks, but their imitation-learning-based policies remain brittle in precise physical interactions due to compounding execution errors; Can a reinforcement learning policy trained purely in simulation improve the robustness of real-world VLAs zero-shot?
Jack B. Jedlicki, Tanguy Dieudonn\'e, Heng Yang
Long-horizon manipulation often requires reasoning about state absent from the current view, such as a vanished object's location, temporal identity, or the contents of a shuffled container.
Pranay Dugar, Mohitvishnu S. Gadde, Rajesh Mangannavar et al.
Recent progress in learned whole-body controllers for humanoid robots has enabled robust free-space motion by tracking kinematic targets such as joint or end-effector trajectories.
Ting Mao, Yanming Shao, Ziheng Wang et al.
Monocular videos of human manipulation provide abundant dexterous demonstrations, yet reconstructing hand-object interaction from a single view and transferring it to robot hands remain difficult, limiting their direct use for robot execution.
Marwah Basuhai, Tingcong Liu, Ibrahim Alsarraj et al.
Soft continuum robots can exploit distributed compliance for whole-body manipulation, but synthesizing behavior through changing contacts remains difficult. We address whole-body grasping and pick-and-throw from an initially ungrasped state through outcome-based actuation-space optimization.
Sai Krishna Reddy Sathi, Anuj Tiwari
Adapting language models to new domains via continual pre-training raises a basic evaluation problem: if the training corpus overlaps with what the model already knows, performance gains cannot be cleanly attributed to new learning rather than pre-existing knowledge.
Chang Liu, Sibo Tian, Litian Gong et al.
Robotic disassembly of end-of-life (EoL) products requires robots to accommodate product variation, follow prerequisite sequences, and approach components precisely for contact-rich manipulation.
Xiaomeng Xu, Han Zhang, Zhian Ruan et al.
Humanoid robots have many degrees of freedom and contact surfaces, yet most loco-manipulation research restricts manipulation to the hands and avoids other body contacts with the environment.
Zirui Wu, Chen Guo, Tianjian Jiang et al.
We introduce P$^{3}$-WAM, a unified world action model for egocentric dexterous manipulation that learns 3D perception, action-conditioned prediction, and goal-directed planning through three complementary modes: Perceive, which reconstructs articulated hand motion and camera trajectories from video; Predict, which generates visual futures under prescribed actions; and Plan, which jointly generates action trajectories and their anticipated visual consequences conditioned on an initial image, a language instruction, and an optional visual goal.
Sergey Arkhangelskiy
Real-world evaluation of vision-language-action (VLA) policies still rests on binary success rate at a fixed timeout with N <= 25 rollouts per condition, almost always without confidence intervals or
Wenqing Tian, Zeyu Zhang, Zhaocheng Liu et al.
Astra can act, yet reliable manipulation depends on the system through which it observes and controls the world. We introduce PhysEvo, a framework for physical recursive self-improvement (RSI) around a single frozen model.
Sang Min Kim, Jinwoo Seo, Hyeongjun Heo et al.
We present POIL, a point-based one-shot imitation learning framework with stable dynamical systems.
Jonghoon Lee, Seong Hyeon Park, Byungwoo Jeon et al.
Vision-language-action (VLA) policies have shown strong potential for general-purpose manipulation, yet they often fail on novel, out-of-distribution objects whose appearance or geometry deviates from the training distribution.
Jonghoon Lee, Seong Hyeon Park, Byungwoo Jeon et al.
Vision-language-action (VLA) policies have shown strong potential for general-purpose manipulation, yet they often fail on novel, out-of-distribution objects whose appearance or geometry deviates from the training distribution.
Ruiqi Wang, Dezhong Zhao, Xuesu Xiao et al.
PrefO2O aims to remove the per-task cost of preference-based RL by reusing human preferences from related source tasks, so a new target task needs no human preference labels.
Léonie Plancoulaine, Sylvain Guégan, Franck Plestan et al.
Underactuated robotic hands achieve adaptive and robust grasping with a reduced number of actuators, but predicting the stable equilibrium pose of the grasped object remains a significant challenge.
Qiuyue Wang, Mingsheng Li, Jian Guan et al.
Embodied intelligence is often studied through specialized models for individual tasks such as manipulation or navigation, resulting in fragmented capabilities and limited generalization across tasks,
Kai Stewart, Yasunori Toshimitsu, Robert K. Katzschmann
Dexterous in-hand manipulation of a grasped object with an anthropomorphic hand is an unsolved frontier for robot dexterity. The contact-richness and highly dynamic nature of object-hand interactions tend to require extensive modeling or data-collection efforts for learning-based approaches.
Yuyao Liu, Jiayuan Mao, David Hsu et al.
Coding agents have demonstrated enormous success in solving complex programming problems.
Sang Min Kim, Alexander Alexiev, Tzu-Yuan Lin et al.
Robust dexterous grasping requires maintaining physical stability despite contacts interactively evolving across the entire hand. A precomputed force distribution can easily fail under object motion, modeling errors, or external disturbances.
Yen-Jen Wang, Haozhe Jiang, Shuying Deng et al.
Building reliable robot capabilities across diverse tasks requires substantial human effort to develop and maintain skills, design rewards, and integrate perception with control.
Yuxuan Chen, Wanruo Zhang, Xiao Li
Vision-Language-Action (VLA) models have recently achieved promising performance in robotic manipulation. However, existing benchmarks mainly evaluate generalization on static manipulation tasks and largely overlook dynamic interaction scenarios.
Chunyu Jiang, Yichen Cai, Shuai Li et al.
Contact-rich manipulation requires a robot to be precise in free space to establish the intended contacts, and compliant once in contact to avoid excessive forces.
Linan Deng, Xing Liu, Lin Hong et al.
Dexterous in-hand manipulation requires coordinated control of multiple actuated joints, and a runtime joint fault can abruptly disrupt the contact configuration required for successful manipulation.
Shukai Gong, Xuanran Zhai, Yintianrun Zhang et al.
Long-horizon compositional manipulation has become increasingly important for real-world robot deployment, where a single task involves multiple coordinated subtasks.
Shrutheesh R. Iyer, Thomas Cohn, Zachary Kingston
Robots often must satisfy one or more constraints during motion planning for real-world tasks. When such constraints reduce the valid configuration space to a measure-zero subset, sampling based planning algorithms require modifications to draw feasible samples.
Minhyeong Kang, Sanghyun Kim
This paper presents a motion planning framework that unifies equality and inequality constraints within a single geometric formulation for sampling-based planning in high-dimensional robotic systems.
Bince Qu, Wei Chen, Bo Zhang
The rise of large language models (LLMs) has opened a path toward generalist embodied AI, raising the question of how LLMs can act in the physical world.
Huashuo Lei, Wenxuan Song, Huarui Zhang et al.
Memory is a critical component of robotic intelligence, as robots must rely on past observations and actions to accomplish long-horizon tasks in partially observable environments.
Han Qi, Heng Yang
Robot manipulation policies must generalize across visual shifts while preserving scene context relevant to action.
Zhenxuan Fan, Bo Zhang, Yutong Lin et al.
Vision-Language-Action (VLA) models have shown promising progress in language-conditioned robotic manipulation.
Sida He, Lingxi Xie, Yunning Cao et al.
General-purpose multimodal agents can write robot-control programs, but repeated exploration and model-mediated action selection can make execution slow.
Chaoyu Zhang, Guangshen Ma, Mark Draelos
In medical robotics, data-driven soft-tissue shape control requires deforming tissue toward desired morphologies through sequential physical interactions.
Yunfan Jiang, Yevgen Chebotar, Ruijie Zheng et al.
Recent robot foundation models operate with single-step or short-history visuomotor context.
Yinghua Zhou, Junjie Ye, Yiqi Zhao et al.
World Action Models (WAMs) couple action generation with future visual prediction for robotic manipulation. However, completing the joint video-action denoising process at each replanning cycle incurs substantial latency, delaying action updates and limiting closed-loop responsiveness.
Zerun Wang, Vivek Kamat, Shekhar Bhansali
Manipulating fragile objects remains challenging as robots must understand the state of what they grasp, such as slip or fracture, to respond appropriately, especially when material properties are unknown.
Chengyue Huang, Khang Vo Huynh, Sebastian Elbaum et al.
Robotic manipulation is typically evaluated by task success, but successful completion does not guarantee safe execution. Many rl failures are temporal: a robot may touch a clean surface after contamination or release an object before it is fully inside an enclosure.
A K M Nadimul Haque, Sheila Sutjipto, Marc G. Carmichael et al.
Skill adaptation frameworks based on reinforcement learning often require restrictive assumptions to maintain stability, such as fixed observations or tightly controlled exploration schedules.
Yijie Lu, James Zhao, Weiming Zhi
Learning from human videos offers a promising route to acquiring diverse manipulation skills. Extending this capability beyond tabletop settings to long-horizon mobile manipulation requires adapting and composing demonstrated interactions across changing scenes and robot configurations.
Mahed Dadgostar, Guillem Aleny\`a, J\'ulia Borr\`as
Although cloth is known to exhibit different outcomes under repeated fast dynamic motions, even when the same trajectory is applied, this variability has not yet been systematically characterized.
Shutong Ding, Zejia Zhong, Zhongyi Wang et al.
Recent advances in reinforcement learning (RL) have achieved great successes by leveraging the multimodality and exploration capability of diffusion policies. Among these approaches, one representativ
Yu Qi, Zhang Ye, Xinyi Xu et al.
Compositional generalization is essential for robot to follow diverse instructions. However, pretrained policies are known to take shortcuts, deferring to salient cues rather than grounding language.
Runhan Huang, Xiangcheng Zhang, Haoyu Chen et al.
Robot learning has advanced by scaling demonstration datasets and model capacity; however, deploying a new robot embodiment or task typically necessitates a dedicated teleoperation corpus.
Yishu Li, Liyuan Geng, Xinyi Mao et al.
While pretrained robotic policies exhibit impressive capabilities in controlled environments, unobserved physical properties and dynamics require these policies to rapidly adapt during deployment.
Tengbo Yu, Jiahao Wu, Daohan Li et al.
Imitation learning for dexterous hands is bottlenecked by the difficulty of collecting contact-rich demonstrations that transfer faithfully to the robot.
Jinshi Zhao, Libaihe Tian, Alex Ranne et al.
Long, slender instruments are essential for minimally invasive surgery (MIS), but achieving precise distal manipulation becomes increasingly difficult when actuation extends along the entire instrument.
Hongxin Zhang, Chunru Lin, Tsun-Hsuan Wang et al.
While Vision-Language-Action (VLA) models demonstrate impressive capabilities in robotic manipulation, their memoryless nature renders them brittle to test-time environment shifts, particularly hardware shifts caused by wear or imperfect calibration.
Shaekh Mohammad Shithil, Lantao Liu
SEVI targets unsafe interactions during online reinforcement learning on physical robots by deciding when a proposed action should be replaced with a recovery behavior.
Wen Sin Lor, Jun Wang, Suyi Li
Many physical properties relevant to robotic manipulation are hidden from vision. A sealed object, for example, may reveal little about its center of mass (COM) or internal contents until it is lifted, shaken, or otherwise dynamically perturbed.
Yibo Liu, Stanko Oparnica, Simon Shewchun-Jakaitis et al.
Contact-rich manipulation is fundamental in robotics but poses significant challenges due to uncertainties in relative poses, such as misalignments and small clearances in peg-in-hole tasks.
Al Bashir, Partho Ghose, Yibin Wang et al.
Robotic harvesting requires precise manipulation to ensure accurate positioning of the end-effector, especially in a constraint workspace.
Juyi Sheng, Hua Wang, Mengyuan Liu
World action models (WAMs) combine robot action generation with future state prediction. Existing WAMs typically predict videos or learned visual latents, which represent interaction geometry only implicitly and may retain appearance information unrelated to control.
He Zhu, Lusen Zhao, Kwan Man Cheng et al.
Large-scale demonstrations have driven unprecedented progress in robot learning, yet collecting robot data through teleoperation is expensive and difficult to scale to diverse environments and long-horizon tasks.
Sofya Konstantinova, Miguel Altamirano Cabrera, Artem Lykov et al.
Vision-language-action (VLA) policies enable diverse robotic manipulation but can fail during execution without recognizing their own errors. Human observers provide complementary signals, as unexpected robot behavior can trigger rapid vocal, facial, or verbal reactions before failure is completed.
Jingyun Yang, Baiyu Shi, Timothy Yu et al.
A growing body of work suggests that tactile sensing gives robot policies contact information that complements vision in dexterous manipulation. However, visuo-tactile robot data remains scarce: dexterous demonstrations require teleoperating robots, which limits dataset scale.
Tianxing Chen, Wenwei Lin, Chuqiao Lyu et al.
Generalist robot policies benefit from large-scale visual pretraining but often struggle with contact-rich dexterous manipulation, where force, compliance, and incipient slip are hard to observe from images.
Pengteng Li, Weiyu Guo, He Zhang et al.
SOMA introduces an explicit spatial memory framework that enables VLA models to track and manipulate objects even after they leave the camera's field of view.
Jiayu Wang, Yue Yu, Bin Zhu et al.
Frontier multimodal foundation models (e.g., GPT-6 Astra) have recently shown strong potential for direct robotic control, yet their performance on fine manipulation remains limited.
Rhythm Syed, Jean Mercat, Sedrick Keh et al.
Vision-language-action models (VLAs) inherit strong semantic grounding from pretrained vision-language backbones but are typically optimized for predicting actions rather than future observations. They can see and act, but they do not imagine the future before acting.
Mousumi Das, Aditeya Prajapati, Abrar Anwar et al.
Robot manipulation systems using Vision-Language-Action (VLA) model backbones typically use just one VLA for task execution. However, individual VLAs do not perform well across different task states and environments.
Tingcong Liu, Aye Phyu Phyu Aung, Junjie Xiong et al.
Soft and continuum robots enable manipulation through distributed body deformation and contact, yet translating language and visual context into executable whole-body actuation remains a fundamental challenge.
Loizos Hadjiloizou, Rodrigo Pérez-Dattari, Noémie Jaquier
Formalizes cross-space symmetry compositions in robotics and derives equivariant network architectures that exploit multiple simultaneous symmetries for improved generalization.
Yuto Tanaka, Kyo Kutsuzawa, Martina Doku et al.
Imitation learning enables robots to acquire complex skills directly from massive demonstration datasets, but its performance degrades severely when datasets are contaminated with suboptimal or noisy demonstrations.
Ariel Herrera, Xueyang Kang, Atal Anil Kumar
Robotic manipulation of textiles remains challenging because continuous deformation and self-occlusions hinder the robust visual perception required to estimate the cloth's state.
Arunim Joarder, Arjun Bhardwaj, René Zurbrügg et al.
Tactile sensing provides direct measurements of contact interactions that are essential for robotic manipulation.
Arunim Joarder, Arjun Bhardwaj, René Zurbrügg et al.
Tactile sensing provides direct measurements of contact interactions that are essential for robotic manipulation.
Zhaofeng Luo, Xinyu Lu, jaehoon Choi et al.
Applying adhesive tape to secure wire harnesses or seal packages requires robots to coordinate a flexible strip, a moving roll, and surfaces that attach and detach.
Nithin S. Kumar, Joshua Gaston, D. Caleb Rucker et al.
Soft actuators enable dexterous and compliant interaction, but closed-loop task-space control remains challenging due to strong nonlinearities, distributed deformation, and uncertainty in their dynamics.
Niloufar Amiri, Houman Masnavi, Farrokh Janabi-Sharifi
This paper investigates temporal neural networks for \mbox{end-effector} position \mbox{estimation} of an aerial continuum manipulator (ACM) operating under aerodynamic effects induced by the unmanned aerial vehicle (UAV).
Mahmud Hasan Saikot, Sydney Spiegel, Sudheera Akalanka Kariyawasam et al.
Robots that can change their morphologies and behaviors for different tasks and environments hold great promise for adaptable, multifunctional systems.
Som Sagar, Ransalu Senanayake
Robot manipulation policies are usually trained under the assumption that a commanded action produces the same motion as it did during training even after hours of operation.
Jin Xu, Yu-Ping Chen, Ayanna Howard
This paper presents THRIVE (Therapeutic Humanoid Robot In Virtual Environment), an at-home rehabilitation platform that integrates a suite of virtual-reality upper-body rehabilitation games, a real-time camera-based motion-tracking system, and a socially interactive robot therapist.
Matthew M. Hong, Jesse Zhang, Anusha Nagabandi et al.
Fine-tuning pre-trained robot policies with reinforcement learning (RL) often inherits the bottlenecks introduced by pre-training with behavioral cloning (BC), which produces narrow action distributions that lack the coverage necessary for downstream exploration.
Vivek Chavan, Yahuan Shi, Oliver Heimann et al.
Vision-language-action (VLA) models can execute short manipulation skills, but remain brittle in long-horizon procedures requiring persistent task state, dependency-aware reasoning, conditional decisions, and reliable grounding.
Masato Kobayashi, Takeru Tsunoori
Underwater visuomotor policy learning has focused primarily on single manipulators, while bimanual imitation learning has been studied largely in air.
Ralf Römer, Maximilian Seeliger, Saida Liu et al.
Quantifies epistemic uncertainty in flow-matching VLAs using velocity-field disagreement (VFD) across a small ensemble, enabling failure detection at deployment and sample-efficient active fine-tuning (SAVE).
Wei Li, Rui Shao, Jie He et al.
Recent advances in Vision-Language-Action (VLA) models have improved robotic manipulation, yet observation-to-action learning remains limited by a fundamental transition realizability gap, manifested in three tightly coupled problems: (i) Transition ambiguity.
Wenye Yu, Yijin Chen, Kaiyi Xiong et al.
Real-world post-training of generalist robot policies faces two challenges: partial observability makes action selection ambiguous, while offline post-training methods delay feedback from deployment experience.
Boyao Han, Chen Shi, Jingjing Qian et al.
Vision-Language-Action (VLA) models have emerged as powerful foundations for robotic manipulation, but their reliance on fixed camera configurations during training makes them brittle to changes in camera count or pose during deployment.
Xincheng Tang, Yiji Chen, Youhan Xie et al.
Door opening and traversal is a long-horizon loco-manipulation task that requires precise handle interaction and coordinated base-arm control. We present Video2DoorTraversal, a single-video real-to-sim-to-real framework for wheel-legged mobile manipulators.
Hisham Khalil, Neil Fernandes, Thomas M. Kwok et al.
Contact-rich manipulation requires precise tracking and mechanical compliance, where variable impedance control can improve robustness in task success, whereas static compliance cannot adapt to varying contact constraints.
Mert Albaba, Jens Beißwenger, Anna Manasyan et al.
Teaching a humanoid to follow instructions with its whole body runs into two obstacles. Its action space is large and tightly coupled: legs, arms, and fingers must move together while the robot keeps its balance, which makes joint-level actions hard to learn.
Jeon Ho Kang, Igal Tamarkin, Ethan Niu et al.
Real-world robotic disassembly requires long-horizon execution, where robots must perform ordered sequences of manipulation tasks across multiple parts within a single scene.
Chang Song, Bin Qian, Yan Feng et al.
Long-horizon robot manipulation with Vision-Language-Action (VLA) policies remains vulnerable to execution-time deviations, as final task success provides little information for diagnosing and correcting failures caused by action noise, object displacement, or goal misalignment.
Suyu Ye, Zheyuan Zhang, Vaishnav Tadiparthi et al.
Robots operating in the physical world will increasingly need to coordinate with other robots, particularly in manipulation tasks where an object may be too large or heavy for a single robot to carry alone.
Vivek Chavan, Pengtao Xie, Yahuan Shi et al.
Visuomotor imitation policies can achieve high performance under in-distribution visual conditions yet fail when visually similar objects or receptacles are introduced.
Jiaming Wang (National University of Singapore)
Can a robot improve itself the way coding agents now improve software? We built an agentic system to find out.
Hu, Songbo, Liao, Qiayuan, Chi, Yufeng et al.
Humanoid robots still struggle to plan contact-rich whole-body manipulation from egocentric RGB and proprioception. Workhorse learns such manipulation from robot-free human demonstrations. A visual planner predicts five-link targets: the poses of the torso, both wrists, and both feet.
Jaehwi Jang, Zhaoyuan Gu, Alfred Cueva et al.
Whole-body humanoid manipulation of bulky, deformable, and shared-load objects requires distributed contact sensing and explicit force regulation, yet most imitation policies treat contact force only implicitly.
Kai Xiong, Hongjie Fang, Lixin Yang et al.
Effectively handling the interplay between spatial perception and action generation remains a critical bottleneck in robotic manipulation.
Zhe Li, Zhenzhe Zhang, Yangyang Wei et al.
Humanoid household tasks often require concurrent loco-manipulation, where the robot must move, adjust posture, maintain balance, and manipulate objects as a single coordinated behavior.
NVIDIA, :, Johan Bjorck et al.
General-purpose robots need a versatile body and an intelligent mind. Recent advancements in humanoid robots have shown great promise as a hardware platform for building generalist autonomy in the human world.
Haoran Jiang, Jin Chen, Qingwen Bu et al.
Humanoid robots require precise locomotion and dexterous manipulation to perform challenging loco-manipulation tasks. Yet existing approaches, modular or end-to-end, are deficient in manipulation-aware locomotion.
Alexander Khazatsky, Karl Pertsch, Suraj Nair et al.
A 350-hour dataset of diverse real-world robot manipulation across 22 robots in 71 scenes, designed to train scalable and generalist imitation learning policies.
Remi Cadene, Simon Alibert, Alexander Soare et al.
Hugging Face's LeRobot is an open-source PyTorch framework providing pretrained models, datasets, and training scripts for imitation and reinforcement learning on real robots, lowering the entry barrier to robot learning.
Dibya Ghosh, Homer Walke, Karl Pertsch et al.
Octo is a large open-source transformer-based generalist robot policy trained on 800k trajectories, supporting language-conditioned and goal-image-conditioned control across 9 robotic platforms with efficient fine-tuning on consumer GPUs.
Peiqi Liu, Yat Long Lo, Ted Xiao et al.
OK-Robot uses off-the-shelf VLMs (CLIP, OWL-ViT) and LLMs (GPT-4) to perform open-ended object manipulation in unseen homes without any training, achieving 58% success on real-world pick-and-place tasks.
Karl Pertsch, Oliver Groth, Jonas Frey et al.
π0 is a 3.5B-parameter VLA flow model from Physical Intelligence that achieves state-of-the-art general robot manipulation by mixing online RL with high-quality human demonstrations, available as an open-source PyTorch implementation.
Moo Jin Kim, Karl Pertsch, Siddharth Karamcheti et al.
A 7B-parameter open-source Vision-Language-Action model pre-trained on 970k real-world robot demonstrations, achieving strong generalization across robots and tasks.
Anthony Brohan, Noah Brown, Justice Carbajal et al.
Google DeepMind's VLA model combining a vision-language foundation model with robot action outputs, showing emergent generalization to novel objects, backgrounds, and semantic instructions far beyond training data.
Xuanlin Li, Kyle Hsu, Jiayuan Gu et al.
A benchmark environment for evaluating real-world robot manipulation policies in simulation using real-world video prompts, enabling zero-shot sim-to-real transfer evaluation without physical hardware.
Tony Z. Zhao, Vikash Kumar, Sergey Levine et al.
ACT employs a transformer-based action chunking policy and temporal ensembling to perform precise bimanual manipulation on a $5k ALOHA robot setup.
Cheng Chi, Siyuan Feng, Yilun Du et al.
A behavior cloning approach that models robot policies as conditional diffusion processes, enabling multimodal action distributions and smooth action sequences.
Charlie C. Kemp, Bilsen D. Kim, Blake S. Kim et al.
A low-cost mobile manipulator designed for indoor home and office environments, with a large open-source ecosystem and teleoperation-based data collection.
Jiayuan Gu, Sean Xiang, Stone Tao et al.
ManiSkill is a GPU-parallelized robotics simulation benchmark with 20+ manipulation tasks, PartNet-Mobility assets, and unified observation/action spaces, supporting RL, IL, and VLA training at millions of steps per hour.
Open X-Embodiment Collaboration
The largest collaborative robot learning dataset ever assembled, spanning 22 robots from 21 institutions, enabling training of generalist RT-X policies that exhibit positive cross-embodiment transfer.
Anthony Brohan, Yevgen Chebotar, Chelsea Finn et al.
RT-1 is a 35M-parameter transformer trained on 130K robot demonstrations that generalizes to new tasks, objects, and environments, forming the foundation for Google's RT-2 and RT-X line of VLA models.
Wenlong Huang, Chen Wang, Ruohan Zhang et al.
VoxPoser uses LLMs to compose 3D value maps from free-form language instructions, enabling zero-shot closed-loop manipulation planning with model-based trajectory synthesis over dynamic environments.
Mohit Shridhar, Lucas Manuelli, Dieter Fox
CLIPort fuses CLIP's semantic understanding with Transporter Networks' spatial precision to perform language-conditioned manipulation tasks like 'stack the red block on the blue block' with 2D pick-and-place affordances.
Mike Lambeta, Po-Wei Chou, Stephen Tian et al.
An affordable, compact, high-resolution vision-based tactile sensor optimized for in-hand manipulation with real-time fingertip deformation sensing.
Josh Tobin, Rachel Fong, Alex Ray et al.
Introduced domain randomization for vision-based manipulation: randomize simulation textures, lighting, and camera parameters so that the real world looks like just another variation.
Read these papers in order to build expertise in Manipulation.
Jiaxin Shi, Xidong Zhang et al. · 2026
Octi Zhang, Mateo Guaman Castro et al. · 2026
Mozes Jacobs, Aiden Swann et al. · 2026
Chen Cai, Steven Liu · 2026
Mathilde Kappel, Clémence Grislain et al. · 2026
Dang Viet Anh Nguyen, Rajesh Piplani et al. · 2026
…and 236 more papers.
