JoyAI-RA 0.1: A Foundation Model for Robotic Autonomy
Abstract
Robotic autonomy in open-world environments is fundamentally limited by insufficient data diversity and poor cross-embodiment generalization. Existing robotic datasets are often limited in scale and task coverage, while relatively large differences across robot embodiments impede effective behavior knowledge transfer. To address these challenges, we propose JoyAI-RA, a vision-language-action (VLA) embodied foundation model tailored for generalizable robotic manipulation. JoyAI-RA presents a multi-source multi-level pretraining framework that integrates web data, large-scale egocentric human manipulation videos, simulation-generated trajectories, and real-robot data. Through training on heterogeneous multi-source data with explicit action-space unification, JoyAI-RA effectively bridges embodiment gaps, particularly between human manipulation and robotic control, thereby enhancing cross-embodiment behavior learning. JoyAI-RA outperforms state-of-the-art methods in both simulation and real-world benchmarks, especially on diverse tasks with generalization demands.
Cite
@article{arxiv.2604.20100,
title = {JoyAI-RA 0.1: A Foundation Model for Robotic Autonomy},
author = {Tianle Zhang and Zhihao Yuan and Dafeng Chi and Peidong Liu and Dongwei Li and Kejun Hu and Likui Zhang and Junnan Nie and Ziming Wei and Zengjue Chen and Yili Tang and Jiayi Li and Zhiyuan Xiang and Mingyang Li and Tianci Luo and Hanwen Wan and Ao Li and Linbo Zhai and Zhihao Zhan and Xiaodong Bai and Jiakun Cai and Peng Cao and Kangliang Chen and Siang Chen and Yixiang Dai and Shuai Di and Yicheng Gong and Chenguang Gui and Yucheng Guo and Peng Hao and Qingrong He and Haoyang Huang and Kunrui Huang and Zhixuan Huang and Shibo Jin and Yixiang Jin and Anson Li and Dongjiang Li and Jiawei Li and Ruodai Li and Yihang Li and Yuzhen Li and Jiaming Liang and Fangsheng Liu and Jing Long and Mingxi Luo and Xing Pan and Hui Shen and Xiaomeng Tian and Daming Wang and Song Wang and Junwu Xiong and Hang Xu and Wanting Xu and Zhengcheng Yu and He Zhang and Jiyao Zhang and Lin Zhao and Chen Zhou and Nan Duan and Yuzheng Zhuang and Liang Lin},
journal= {arXiv preprint arXiv:2604.20100},
year = {2026}
}