Leveraging Generative AI Prompt Programming for Human-Robot Collaborative Assembly

Abstract

In manufacturing, traditional robotic programming methodologies have often been focused on independent operation, offering limited capabilities for seamless human-robot collaboration. This paper introduces a paradigm shift in collaborative production systems by leveraging generative artificial intelligence (AI), specifically large language models (LLMs). Contrary to traditional methods that rely on pre-defined assembly instructions, this paper introduces a novel framework employing primitive knowledge of the production process, including product design and required assembly steps. By integrating LLMs and a behavior tree-based system control, this approach enables programmers to rapidly deploy collaborative assembly procedures by expediting the programming of robotic operations. The system also incorporates Natural Language Processing (NLP) technologies, which facilitate real-time alterations in assembly steps, leading to reduced overall production time. The framework’s behavior tree-based control architecture allows for dynamic adaptability, offering optimized solutions across a range of assembly scenarios. The results of the framework’s deployment suggest that this innovative programming paradigm significantly enhances both the adaptability and efficiency of collaborative manufacturing settings.