Where evolvable production systems meet complexity science

Modern control approaches are either designated as complex or tackle complex systems. The concept of Evolvable Production System (EPS) entails an iterative system design/control loop that supports system Adaptation and Evolution. Both concepts are borrowed from natural systems and, if properly instantiated and explored for the sake of production systems, may come to include a set of self-organizing and emergent process. One of the main challenges in EPS research is to frame these developing natural concepts and set them in a Mechatronic Agent-based framework. Such an approach also brings into the equation the possibility to handle and, potentially, exploit emergent behaviour. Inevitably, the coping of dynamically evolving setups with unforecasted behaviours raises the complexity issue quite substantially, hence the need to begin to analyse the EPS systems as Complex Systems. This paper provides an introduction to the context of Complex Systems, in particular about the self-organization and emergence components, and, from this integration, details how the bridging of the technological and functional challenges is being adopted within EPS. The work is based on assembly systems built within the framework of several European projects, including EUPASS, IPROMS and IDEAS. Finally, the article will provide with an initial evaluation and critical review of the complexity approach.