A two stage algorithm to design and assess product family

This paper presents a two-stage algorithm to design product families and allocate products to assembly lines. Product family design is triggered by the existence of large variety of products in a typical company portfolio. The literature shows that physical similarity and process similarity are the most important factors in product family design. The objective of this paper is to classify products based on their physical appearance such that similar products can be manufactured in the same assembly line. The proposed model starts with determination of similarity ratio among products using Jaccard index. Afterwards, an algorithm is developed for the product allocation to the assembly lines in a way that each assembly line indicates a product family. Additionally, weighted subjective product score is calculated. Cluster analysis is used and products are clustered using weighted subjective product score. Comparison of the cluster analysis and proposed model using Jaccard index are presented and discussed. It is concluded that classification of the products using Jaccard index improves overall efficiency of product lines and gives an efficient preliminary rough cut capacity planning.