Chapter 7 - Deployment roadmap of proactive human–robot collaboration

This chapter presents a stepwise procedure for the development of Proactive HRC systems comprising four key modules: scene perception, knowledge representation, decision making, and collaborative control. For each module, we provide a comprehensive research roadmap of related technologies and offer an advanced algorithm as a feasible solution. The perception module is dedicated to perceiving the human–robot–workspace environment, as detailed in Section 7.1. Meanwhile, knowledge representation focuses on acquiring semantic knowledge of manufacturing tasks and transferring human expertise to robots for cognitive inference, as illustrated in Section 7.2. In Section 7.3, we delve into the decision-making module, which empowers the HRC system to make intelligent decisions for optimized trajectory planning and human information support, adapting to changing environmental conditions. Additionally, Section 7.4 provides an overview of various algorithms for robot collaborative control at the operational level. These four aspects have witnessed the widespread adoption of cutting-edge cognitive computing techniques such as deep learning, reinforcement learning, transfer learning, large language model, etc., resulting in significant enhancements to Proactive HRC system performance.