Cyber-Physical Production Systems (CPPS) are inherently complex due to their tightly integrated components and dynamic interactions. This complexity is further intensified by disturbances, which can propagate across system components, disrupt workflows, and degrade performance. To manage such disturbances, it is essential to consider disturbance handling as a core operational scenario from the earliest stages of control architecture design. While various paradigms exist for distributed control, there remains a lack of conceptual frameworks that offer generic and integrative design patterns for adaptive control architectures. This article proposes the human Biological Immune System (BIS) as a design pattern to guide the development of knowledge-based, distributed control architectures for CPPS. The contribution is two-fold. First, an immune-inspired adaptive control meta-model is developed using ArchiMate enterprise architecture modeling. This meta-model provides a reusable pattern for designing adaptive control systems. Second, the meta-model is instantiated in a Product-Driven Control (PDC) architecture, where smart products are equipped with embedded agents capable of detecting, evaluating, and reacting to disturbances autonomously. The implementation is validated through a multi-agent simulation of an industrial-scale learning factory, where product agents use local decision-making mechanisms—including similarity-based case retrieval, multi-criteria coordination, and immune-inspired optimization—to adapt to disturbances. Simulation results demonstrate the architecture’s ability to improve responsiveness, reduce downtime, and maintain performance under varying disturbance scenarios, offering promising directions for scalable and resilient CPPS design.
Darmoul et al. (Sun,) studied this question.