Evolution of a self-organizing manufacturing network with homophily and heterophily

Design, manufacturing, and aftersales service are the three major stages in a product's lifecycle. The implementation of product-service systems has integrated the whole value chain and stimulated the potentials in improving the product design according to the operational data collected from end users. However, it is still difficult to improve the resource configuration at the manufacturing stage based on the analysis of such data, especially in the era of smart manufacturing, where the manual process planning is mostly replaced by self-organizing mechanisms. This preliminary research investigated the evolution of self-organizing manufacturing networks from the angle of evolutionary dynamics in complex networks. First, a common measure of the resource capacity was mathematically modeled to calculate the similarities among resources. Then, an iterative process was designed to improve the structure of the self-organizing network, considering the internal characteristics (e.g., the homophily and heterophily) of and the external evaluation (e.g., the operational data and users' feedback) to the manufacturing resources. Numerical experiments showed that the manufacturing resource network remained flexible and tended to form functional clusters, which can possibly increase the system robustness and improve the quality of products.