SELF-VLA: A Skill Enhanced Agentic Vision-Language-Action Framework for Contact-Rich Disassembly

TL;DR

SELF-VLA introduces a skill-enhanced vision-language-action framework that significantly improves robotic disassembly performance, addressing variability and complexity in contact-rich, long-horizon tasks, and surpasses existing models in accuracy and adaptability.

Contribution

The paper presents a novel agentic VLA framework with integrated disassembly skills, enhancing generalization and performance in complex robotic disassembly tasks.

Findings

Outperforms state-of-the-art VLA models on contact-rich tasks

Demonstrates improved generalization to variable EoL products

Achieves higher success rates in complex disassembly scenarios

Abstract

Disassembly automation has long been pursued to address the growing demand for efficient and proper recovery of valuable components from the end-of-life (EoL) electronic products. Existing approaches have demonstrated promising and regimented performance by decomposing the disassembly process into different subtasks. However, each subtask typically requires extensive data preparation, model training, and system management. Moreover, these approaches are often task- and component-specific, making them poorly suited to handle the variability and uncertainty of EoL products and limiting their generalization capabilities. All these factors restrict the practical deployment of current robotic disassembly systems and leave them highly reliant on human labor. With the recent development of foundation models in robotics, vision-language-action (VLA) models have shown impressive performance on standard robotic manipulation tasks, but their applicability to complex, contact-rich, and long-horizon industrial practices like disassembly, which requires sequential and precise manipulation, remains limited. To address this challenge, we propose SELF-VLA, an agentic VLA framework that integrates explicit disassembly skills. Experimental studies demonstrate that our framework significantly outperforms current state-of-the-art end-to-end VLA models on two contact-rich disassembly tasks. The video illustration can be found via https://zh.engr.tamu.edu/wp-content/uploads/sites/310/2026/03/IROS-VLA-Video.mp4.