A method for human–robot complementary collaborative assembly based on knowledge graph

A method for human–robot complementary collaborative assembly based on knowledge graph

연구 분야: Databases

학회: CCF Transactions on Pervasive Computing and Interaction

Human–robot collaboration plays a key role in improving the efficiency of assembly and disassembly tasks. However, given the increasing complexity of tasks, identifying methods to enhance efficiency and safety in assembly and disassembly operations remains a pressing challenge. In this paper, we propose a knowledge graph-based human–robot collaborative assembly approach (KGCCA) to address this challenge. We employ the constructed knowledge graph to intelligently allocate assembly tasks and foster complementary collaboration between humans and robots, addressing the challenges posed by increasingly complex task requirements. Our main innovations include utilizing the knowledge graph for guiding intelligent planning decisions, proposing an enhanced PageRank (PR)-based algorithm to improve task allocation efficiency, and devising strategies for task disassembly to facilitate efficient, flexible, and safe human–robot collaboration. Through experimental validation, it has been demonstrated that our knowledge graph-based human–robot collaboration algorithm effectively allocates sub-tasks, enabling humans and robots to collaborate efficiently and accurately in completing the block-building task. These results underscore the proposed algorithm’s significant research and practical value in the domain of human–robot cooperative assembly.