This work organizes the embodied data ecosystem as a pyramidspanning five complementary sources: real-robot data, UMI-style data, egocentric and exocentric data, simulation data, and general vision-language data, and further characterize each source in terms of data quality, diversity, reusability, and physical fidelity.
Abstract
Multimodal foundation models learned to see and to speak by consuming the whole internet. Embodied agents admit no such shortcut, since they require data that couple observations with physical states and actions. These signals can be provided, to varying degrees, by multiple data sources. In this work, we organize the embodied data ecosystem as a"pyramid"spanning five complementary sources: real-robot data, UMI-style data, egocentric and exocentric data, simulation data, and general vision-language data. We organize the pyramid around the tension between scalability and robot alignment, and further characterize each source in terms of data quality, diversity, reusability, and physical fidelity. We then analyze recent embodied foundation models through the lens of their data recipes, examining how different sources are selected, aligned, and mixed during pretraining. For embodied brain models, vision-language-action models, and world-action models alike, we relate data composition to capabilities in perception, reasoning, planning, action generation, and world prediction. We close by discussing six open challenges: building large-scale tactile datasets, collecting failure and recovery data, developing scalable data-collection pipelines, aligning actions across embodiments, leveraging egocentric data for dexterous manipulation, and designing principled data recipes for robot learning. We hope this work paves the foundation for the design of next-generation embodied systems.
Vision-language-action (VLA) models have become a dominant paradigm for generalist embodied agents, demonstrating strong complex and long-horizon task completion in structured settings. Yet it remains an open question whether current VLA systems can benefit from more effective architectural design, scale to substantially larger and more heterogeneous data regimes, and achieve broader generalization across tasks and embodiments. To this end, we present GigaBrain-0.7, an embodied foundation model with substantially improved generalization across diverse robot embodiments. Specifically, GigaBrain-0.7 unifies understanding, prediction, and action through a three-system architecture, scales pretraining to over 37,000 hours of heterogeneous embodied data, and introduces one-stage alignment training that jointly optimizes vision-language understanding and multi-embodiment action generation. Compared with the preceding GigaBrain-0 series and prior state-of-the-art models including $\pi_{0.5}$, GigaBrain-0.7 achieves substantial improvements in foundation zero-shot capabilities, language-conditioned instruction following, and post-training task success rates. In particular, on our in-house Maker H01 platform and mainstream robot embodiments, GigaBrain-0.7 demonstrates strong task adaptability and completion ability across both home and industrial scenarios. All training code and pretrained model weights will be released.
GigaBrain Team, Angen Ye, Axiang Sun et al.· 0 citations
Collecting contact-rich robot experiences at scale remains a major bottleneck for generalizable manipulation. Beyond data quantity, robot learning also requires diverse experiences across embodiments, viewpoints, and scenes. Human egocentric videos provide abundant physical interactions, but each video captures only a narrow slice of experience under a single body, camera trajectory, and environment. We propose AnyWorld, a cross-embodiment world modeling framework that expands a single human interaction into diverse robot-native rollouts without paired human-robot demonstrations. Our model factorizes an interaction into action, camera, and embodiment: action controls capture the motion structure, camera controls specify viewpoint evolution, and the target embodiment context defines the acting body and its interaction geometry. This formulation enables independent recomposition of embodiment, viewpoint, and scene factors, allowing a single model to generate many robot-domain experiences while preserving the underlying dynamics and object interactions. We train the model with large-scale human interaction pretraining followed by mixed-embodiment fine-tuning. Experiments show that our model supports controllable recomposition across embodiments, viewpoints, and scenes, and we further demonstrate that the generated data can improve manipulation performance on the RoboCasa GR1 tabletop benchmark and a real IRON humanoid robot. Beyond aggregate gains, we test whether unpaired human experience can be recomposed into robot-native video-action pairs that target a policy gap. Controlled IRON interventions correct a spurious completion prior and establish language-grounded spatial target selection; an action-only counterfactual intervention fails to learn the latter reliably, showing that both action calibration and visual recomposition are necessary.
Cheng Chen, J. Bai, Jiacheng Wei et al.· 0 citations
Results demonstrate reliable cross-embodiment translation and show that robot data generation can be reframed from a hardware collection problem into a scalable, low-resource knowledge transfer problem.
LingBot-Video is presented, a DiT-based video pretraining paradigm specifically tailored for embodied intelligence, and is contributed as the inaugural large-scale, open-source MoE video foundation model to the community, in a pioneering effort to bridge digital creativity and physical actuation.
Shuailei Ma, Jiaqi Liao, Xinyang Wang et al.· 7 citations· ⚡1
The remarkable progress of Multimodal Large Language Models (MLLMs) has attracted increasing attention to extending them to physical entities like legged robots and robotic arms. This typically requires MLLMs to not only grasp multimodal understanding abilities, but also integrate visual-spatial reasoning and physical interaction capabilities. Nevertheless, existing methods struggle to jointly accommodate these capabilities within a shared multimodal model, due to the substantial differences between vision-language reasoning and low-level physical control. In this paper, we present Visual Embodied Brain-1.5 (VeBrain-1.5), a task-level unified framework that connects multimodal perception and spatial reasoning with robot control through a shared MLLM-compatible decision interface. VeBrain-1.5 reformulates robotic control into common text-based MLLM tasks in the 2D visual space, thereby aligning the supervision format and prediction interface of multimodal understanding, spatial reasoning, and robot control. A modular robotic adapter subsequently translates these high-level textual decisions into embodiment-specific motion commands and executes them in a closed loop. From the data perspective, we further introduce VeBrain-1M, a high-quality instruction dataset encompassing various capabilities of VeBrain-1.5. In VeBrain-1M, we take hundreds of hours to collect, curate and annotate the data, and adopt multimodal chain-of-thought (CoT) to mix the different capabilities into a single conversation. Extensive experiments on 13 multimodal benchmarks, 7 spatial intelligence benchmarks and 4 simulation and real-world robotic benchmarks demonstrate the superior performance of VeBrain-1.5 over existing MLLMs like Qwen2.5-VL. To further eliminate the task ambiguity and enhance the reasoning capability, we adopt an offline RL policy learning with verifiable reward to increase the task success rate. When deployed to two robotic entities including legged robots and robotic arms, VeBrain-1.5 shows strong adaptability, flexibility, and compositional capabilities compared to existing methods.
Ganlin Yang, G. Luo, Ziyang Gong et al.· IEEE Transactions on Pattern...· 0 citations
Robot foundation models (RFMs), including vision-language-action (VLA) policies, are often discussed through a scaling view: more data, larger models, and broader benchmarks should improve generalization. In robotics, however, a model can generalize while work still remains before it can run on a robot with a particular body. The work required differs across methods and target robots, and those differences affect practical deployment. We call the gap between reusable models, representations, or data and their use in execution on the target robot the embodiment gap. This survey examines what can be reused across robot embodiments and what must still be implemented on a new robot. We place existing methods on a two-axis map that shows the type of shared structure and the stage at which adaptation is needed for execution on the target robot. We then examine recent work through three overlapping research directions: sharing semantics and perception, sharing robot data and interfaces, and learning correspondence across embodiments. We also propose a reporting framework for adaptation work that success rate alone does not reveal. The framework identifies the work that should be checked when comparing cross-embodiment learning and highlights work that remains on a new robot and questions for future study.
Y. Domae, Keisuke Shirai, Hanbit Oh et al.· Trans. Mach. Learn. Res.· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
