AXIS: A Growable Community-Driven Data Engine for Scalable Robot Manipulation
2026-07-23 • Robotics
Robotics
AI summaryⓘ
The authors created AXIS, a system that helps collect lots of robot manipulation examples from people using just a web browser, making it easier to gather diverse training data. AXIS also automatically creates and checks new tasks, cleans and improves the demonstration data, and organizes everything for easy training and evaluation. Their dataset includes over 207 tasks and 50,000 demonstration runs. They tested different robot learning models and found that training continually with AXIS data improves robot success rates, especially under challenging conditions like sensor noise or camera changes. This work shows that scalable, community-driven data collection can advance robot learning policies effectively.
robot manipulationteleoperationdata augmentationtask benchmarkingtrajectory smoothingrobot learningvision-language-action modelspretrainingdataset scaling
Authors
Mengfei Zhao, Dihong Huang, Yikai Tang, Peihao Li, Mingxuan Yan, Ruiqi Zhuang, Yanjia Huang, Jie Wang, Hai Zhai, Tony Zhou, Rui Zhang, Zhexi Luo, Yuchen Huang, Jianfei Yang, Jiachen Li
Abstract
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. We present AXIS, a growable community-driven data engine and benchmark for scalable robot learning, which enables browser-based teleoperation for large-scale demonstration collection, automatically generates and validates new manipulation tasks, and transforms community-collected demonstrations into training-ready data through automated success checking, quality filtering, trajectory smoothing, and visual and physics-based augmentation. The AXIS dataset currently contains 207 diverse tasks and 50K+ trajectories. Meanwhile, AXIS organizes data into task snapshots and evaluates policies with a systematic held-out protocol. We compare vision-language-action (VLA) policies under a unified AXIS evaluation suite and analyze scaling behavior across different data volumes. Continual pretraining on AXIS substantially improves the overall success rate of $π_{0.5}$ by 5.8%, outperforms the model pretrained on RoboCasa365 by 37.3%, and exhibits consistent scaling with increasing data volume, with the largest gains observed under layout, sensor-noise, and camera perturbations.