Evolutionary approaches to the design and organization of manufacturing systems

Evolutionary computation has shown an interesting potential in the engineering fields. The optimization strategies used, based on a population of solutions and often inspired from nature, have led an increasing number of researchers to publish articles that address several types of problems encountered in the area of manufacturing systems. In this article, focus is brought on problems related to the design and organization of manufacturing systems. A survey has been conducted from a large number of literature references. From these references, we have identified the following types of problems as most addressed: cellular organization, facility layout, and optimization of the workshop configuration, assembly systems (including assembly line balancing). After a first part recalling the main characteristics of evolutionary algorithms and their potential to handle features frequently encountered in manufacturing problems (such as multiple criteria and constraints), each type of problem identified is studied. In particular, we summarize the way researchers have suggested to encode solutions and the main characteristics of the algorithms used. Several opened research directions in the areas investigated have been identified and are suggested.