OpenHelix: A Short Survey, Empirical Analysis, and Open-Source Dual-System VLA Model for Robotic Manipulation

TL;DR

This paper provides a comprehensive survey and empirical analysis of dual-system VLA models for robotic manipulation, introducing an open-source model to facilitate further research and optimization.

Contribution

It offers the first systematic comparison and empirical evaluation of existing dual-system VLA architectures, along with an open-source model for community use.

Findings

Comparison of structural designs of dual-system architectures

Empirical evaluation of core design elements

Open-source VLA model for further research

Abstract

Dual-system VLA (Vision-Language-Action) architectures have become a hot topic in embodied intelligence research, but there is a lack of sufficient open-source work for further performance analysis and optimization. To address this problem, this paper will summarize and compare the structural designs of existing dual-system architectures, and conduct systematic empirical evaluations on the core design elements of existing dual-system architectures. Ultimately, it will provide a low-cost open-source model for further exploration. Of course, this project will continue to update with more experimental conclusions and open-source models with improved performance for everyone to choose from. Project page: https://openhelix-robot.github.io/.