To remain competitive in the midst of global competition and rapidly changing consumer tastes, manufacturers increased the amount of product variety they offer to the market and their responsiveness to changing market needs. Product families and reconfigurable manufacturing systems (RMS) enable manufacturers to cost effectively supply high product variety and to be responsive. However, there is a lack of systematic methods for the joint design of product families and RMSs.

Co-evolution of product families and assembly systems is introduced as a new method for the joint design and reconfiguration of product families and assembly systems within and across product generations. There are two main phases in the co-evolution methodology. The first phase involves the joint design of a product family and assembly system in the first generation. The second phase involves the joint evolution of the product family and assembly system between product generations.

For co-evolution, models are required for: (i) the joint design of a product family and its corresponding assembly system, (ii) the evolution of the product family within the constraints of an assisting assembly system, and (iii) the reconfiguration of the assembly system. Methodologies are introduced for the first and third problems in this dissertation.

Two non-linear integer programming (INLP) formulations are developed for the problem of the concurrent design of a single generation of a product family and assembly system. The objective of the first formulation is to maximize the efficiency of the assembly system and minimize the oversupply of functionality. The objective of the second formulation is to maximize profits. Genetic algorithms are introduced for the solution of these problems.

The assembly system reconfiguration planning (ASRP) problem is also formulated as INLP. Genetic algorithm and dynamic programming procedures are introduced for the solution of this problem. In addition, an algorithm for generating all the possible assembly system configurations is introduced.

Examples are used to demonstrate how the methods for co-evolution introduced in this dissertation can lead to reduced costs and increased responsiveness to market changes.