Date of Award


Publication Type

Doctoral Thesis

Degree Name



Industrial and Manufacturing Systems Engineering


Assembly Line Design, Co-Design of Platforms and Systems, Hierarchical Platform, Phylogenetic Networks, Product Families, Product Platforms


Hoda ElMaraghy




In this dissertation, the main hypothesis is that formation of products families and platforms can be simultaneously achieved with their corresponding assembly lines using a holistic mathematical model to increase the effectiveness of mass customization and decrease development and assembly costs. A Phylogenetic Network algorithm, four different mathematical models, and postponement effectiveness metric have been developed and implemented to prove this hypothesis. The results of this research are applicable to many modular products such as consumer goods such as computers, laptops, tablets, power tools, home appliances and laboratory weighing scales which have multiple variants. The research provides a hybrid approach balancing between platforms production using make-to-stock strategy, then further customization using make-to-order strategy.

The Median-Joining Phylogenetic Network (MJPN) is used to model a delayed differentiation assembly line for a product family. The MJPN is capable of increasing commonality across the product platforms using the Median Vector concept. A Postponement Effectiveness metric was developed and showed that the determined assembly line strategy postponed the products delayed differentiation point more than other found in literature. A Modular Product Multi-Platform Configuration Model is introduced to design optimal products platforms which allow assembly and disassembly of components to form new product variants. A new model of Hierarchic Changeable Modular Product Platforms which defines the optimum hierarchy of the platform components is introduced, to enable delayed product differentiation. A Multi-Period Multi-Platform Configuration Model which accounts for demand fluctuation by including the cost and quantity of inventory of product platforms required for implementing the assembly/disassembly platforms customization was developed. Finally, a global product families and platforms formation mathematical model which fully integrates assembly task assignments, precedence relations, assembly cost was introduced. A family of touch screen tablets was used for illustrating the application and advantages of the newly developed product platform models.

This research makes a number of contributions. This is the first time mathematical models are able to flexibly determine the optimal number of product platforms using customization by assembly and disassembly. Inclusion of hierarchy or assembly sequence in platform formation as a variable is novel. This will eliminate assembly sequence ambiguity when designing platforms with duplicate components. The inclusion of inventory costs and quantities in platform design is also new. Finally, the complete integration of platform formation and assembly line design in one mathematical model is introduced for the first time.