<p>Human beings serve as the primary focus within the realm of intelligent manufacturing, and the development of human-centered intelligent manufacturing represents a significant trajectory for its future advancement. Enhancing human comfort has emerged as a central theme in the study of human-centered intelligent manufacturing. The establishment of safe and reliable human–robot collaboration is fundamental to the realization of this paradigm. The concept of virtual-real interaction, as highlighted by digital twin technology, aligns with the requirements of human–robot collaboration and offers innovative strategies for enhancing human comfort in such interactions. This paper presents a digital twin-driven methodology aimed at improving human comfort in human–robot collaboration. Initially, the mechanisms through which digital twins can enhance human comfort are examined, with a focus on monitoring human work status and evaluating ergonomic working postures. Subsequently, the paper delineates how digital twin-driven monitoring of work status and ergonomic assessments contribute to the enhancement of human comfort. Finally, the paper illustrates these concepts through a case study involving the human–robot collaborative assembly of a mechanical gripper for a handling robot, detailing the application of digital twins to improve human comfort and validating the proposed methodologies.</p>