The Model of Structure Evaluation at the Upstream ‎of Automotive Supply Chains by an Adapted ‎Spectral Clustering

The purpose of this study is to design supply chains' upstream structure evaluation model in the automotive industry with spectral clustering based on the theory of complex adaptive systems. In this research, a method for evaluating the intersectionalities related to the structural complexity (horizontal, vertical, and spatial) of supply chains by considering the functional characteristics of its components based on the resilience paradigm is presented. In this regard, a set of algebraic calculations and computational algorithms have been adapted to evaluate the structural design from the perspective of complex components. In the structural design evaluation model through spectral clustering, it is possible to enter information about supply chains in terms of interactions between components in the form of a network as a comprehensive model called similarity graph. According to the field findings, supply chain characteristics in terms of complexity can have interaction with component processing performance. This means that according to the concept of entanglement, the lack of a favorable environmental structure in supply chains can also negatively affect the resilience performance of its components. Findings from the perspective of achieving a supply chain evaluation model as an integrated whole have provided a suitable practical tool for evaluation and pathology of supply chains from the perspective of risk management.