Towards Generalizable Robotic Manipulation in Dynamic Environments

TL;DR

This paper introduces DOMINO, a large-scale dataset and benchmark for dynamic manipulation, and proposes PUMA, a dynamics-aware VLA architecture that improves generalization and success rates in dynamic environments.

Contribution

The work provides a new dataset, benchmark, and a novel architecture for dynamic manipulation, advancing the capabilities of vision-language-action models in dynamic settings.

Findings

PUMA achieves a 6.3% success rate improvement over baselines.

Training on dynamic data enhances transferability to static tasks.

DOMINO enables systematic evaluation of VLAs in dynamic environments.

Abstract

Vision-Language-Action (VLA) models excel in static manipulation but struggle in dynamic environments with moving targets. This performance gap primarily stems from a scarcity of dynamic manipulation datasets and the reliance of mainstream VLAs on single-frame observations, restricting their spatiotemporal reasoning capabilities. To address this, we introduce DOMINO, a large-scale dataset and benchmark for generalizable dynamic manipulation, featuring 35 tasks with hierarchical complexities, over 110K expert trajectories, and a multi-dimensional evaluation suite. Through comprehensive experiments, we systematically evaluate existing VLAs on dynamic tasks, explore effective training strategies for dynamic awareness, and validate the generalizability of dynamic data. Furthermore, we propose PUMA, a dynamics-aware VLA architecture. By integrating scene-centric historical optical flow and specialized world queries to implicitly forecast object-centric future states, PUMA couples history-aware perception with short-horizon prediction. Results demonstrate that PUMA achieves state-of-the-art performance, yielding a 6.3% absolute improvement in success rate over baselines. Moreover, we show that training on dynamic data fosters robust spatiotemporal representations that transfer to static tasks. All code and data are available at https://github.com/H-EmbodVis/DOMINO.