The manufacturing industry is characterized by large-scale interdependent networks as companies buy goods from one another, but do not control or design the overall flow of materials. The result is a complex emergent structure with which companies connect to each other. The topology of this structure impacts the industry's robustness to disruptions in companies, countries, and regions. In this work, we propose an analysis framework for examining robustness in the manufacturing industry and validate it using an empirical dataset. Focusing on two key angles, suppliers and products, we highlight macroscopic and microscopic characteristics of the network and shed light on vulnerabilities of the system. It is shown that large-scale data on structural interdependencies can be examined with measures based on network science. Keywords Complex network Á Supply Á Manufacturing Á Robustness This article is part of a focus collection on ''Robust Manufacturing Control: Robustness and Resilience in Global Manufacturing Networks''.
Abstract-Inspired by studies in ecological networks, we look for a nested pattern in a large-scale data set describing the global automotive industry, including more than 18 000 firms, their clients, and products. Two bipartite networks are formed, namely, supplier-product distribution and supplier-manufacturer relations. Both networks are found to be significantly nested. The pattern means that suppliers produce proper subsets of what other suppliers produce and rare products are produced only by those suppliers that already produce high numbers of product types. In addition, the manufacturers that procure from few suppliers procure from those that supply to most other manufacturers in the network. Similarly, suppliers that supply to few supply to those manufacturers that procure from most others. A nested structure is more robust than a nonnested structure as disrupted suppliers can be substituted, but nestedness also means that small suppliers face more competition as their production can be redundant. Our finding is contrary to conventional wisdom that associates large diversified firms with efficiency and small specialist firms with rare products, showing that large-scale complex system analysis can lead to the discovery of important systemic characteristics, which are obscured when viewed from local points of view. We then propose a multiagent model that creates more realistic nested structures to study systemic outcomes influenced by topology.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.