In the present HRC, humans and robots operate collaboratively within shared workspaces, focus on ensuring workspace safety through continuous monitoring. However, these existing HRC systems is facing challenges related to the acquisition of high-level task knowledge for bi-directional intention understanding. To surmount these hurdles, this chapter introduces an MR and visual reasoning-based approach. This approach enables mutual-cognitive capabilities, thereby providing cognitive support to both human and robotic agents engaged in various tasks. The mutual-cognitive intelligence lies in several components, including dynamic environment perception, cognitive decision-making processes, and ergonomic robot control. These elements elevate the collaborative capabilities of human–robot teams. To illustrate the practical implications of this approach, the chapter presents a detailed demonstration involving the disassembly of an aging electric vehicle battery. This case study underscores how human and robot operations can be combined to meet the collective needs of the team, thereby enhancing the overall efficiency of task execution.