Date of Award


Degree Type


Degree Name



Industrial and Manufacturing Systems Engineering

First Advisor

ElMaraghy, Hoda


Manufacturing Systems; Mathematical Modeling; Networked Operations Sequence; Product Family Formation; Production Sequencing




This research aims at developing novel methods for utilizing the commonality between part/product variants to make modern manufacturing systems more flexible, adaptable, and agile for dealing with less volume per variant and minimizing total changes in the setup between variants. Four models are developed for use in four important domains of manufacturing systems: production sequencing, product family formation, production flow, and products operations sequences retrieval. In all these domains, capitalizing on commonality between the part/product variants has a pivotal role. For production sequencing; a new policy based on setup similarity between product variants is proposed and its results are compared with a developed mathematical model in a permutation flow shop. The results show the proposed algorithm is capable of finding solutions in less than 0.02 seconds with an average error of 1.2%. For product family formation; a novel operation flow based similarity coefficient is developed for variants having networked structures and integrated with two other similarity coefficients, operation and volume similarity, to provide a more comprehensive similarity coefficient. Grouping variants based on the proposed integrated similarity coefficient improves changeover time and utilization of the system. A sequencing method, as a secondary application of this approach, is also developed. For production flow; a new mixed integer programing (MIP) model is developed to assign operations of a family of product variants to candidate machines and also to select the best place for each machine among the candidate locations. The final sequence of performing operations for each variant having networked structures is also determined. The objective is to minimize the total backtracking distance leading to an improvement in total throughput of the system (7.79% in the case study of three engine blocks). For operations sequences retrieval; two mathematical models and an algorithm are developed to construct a master operation sequence from the information of the existing variants belonging to a family of parts/products. This master operation sequence is used to develop the operation sequences for new variants which are sufficiently similar to existing variants. Using the proposed algorithm decreases time of developing the operations sequences of new variants to the seconds.

Available for download on Thursday, October 26, 2017