In the context of Industry 4.0, numerous technologies are developed as well as new paradigms, causing a rupture with historical production control models. This highlights the needs for new approaches aiming to deploy efficiently these new technologies and paradigms within future industrial systems. On the other hand, human-machine system approaches encourage the cooperation between humans and complex artificial systems to react to unexpected events and to ensure an efficient supervision of these artificial systems. The paper focuses on the design of self-organized production systems cooperating with the humans. A literature review is provided based on two views dealing with such a design: a technical and a human-machine system one. Limits and advantages of both views are presented. A merged view, based on the use of the cognitive work analysis (CWA) approach, is then proposed to ensure an efficient cooperation between the human and a self-organized production system. The proposals will be applied to different systems, namely a cobot, a swarm of autonomous AGV and a set of intelligent products.