Manipulate-Anything: Automating Real-World Robots using Vision-Language Models

TL;DR

Manipulate-Anything introduces a scalable, real-world robotic manipulation data generation method that operates without privileged information or hand-designed skills, outperforming existing approaches and enabling robust policy training.

Contribution

The paper presents a novel method that automates real-world robot data generation without privileged info or pre-designed skills, applicable to any static object.

Findings

Successfully generates trajectories for 7 real-world and 14 simulation tasks.

Outperforms existing methods like VoxPoser in data generation.

Produces demonstrations that train more robust policies than human or other synthetic data.

Abstract

Large-scale endeavors like and widespread community efforts such as Open-X-Embodiment have contributed to growing the scale of robot demonstration data. However, there is still an opportunity to improve the quality, quantity, and diversity of robot demonstration data. Although vision-language models have been shown to automatically generate demonstration data, their utility has been limited to environments with privileged state information, they require hand-designed skills, and are limited to interactions with few object instances. We propose Manipulate-Anything, a scalable automated generation method for real-world robotic manipulation. Unlike prior work, our method can operate in real-world environments without any privileged state information, hand-designed skills, and can manipulate any static object. We evaluate our method using two setups. First, Manipulate-Anything successfully generates trajectories for all 7 real-world and 14 simulation tasks, significantly outperforming existing methods like VoxPoser. Second, Manipulate-Anything's demonstrations can train more robust behavior cloning policies than training with human demonstrations, or from data generated by VoxPoser, Scaling-up, and Code-As-Policies. We believe Manipulate-Anything can be a scalable method for both generating data for robotics and solving novel tasks in a zero-shot setting. Project page: https://robot-ma.github.io/.