Risk Propagation Through a Platform: The Failure Risk Perspective on Platform Sharing

Abstract

A product platform is a set of subsystems and interfaces that are commonly shared by a variety of products. Although platform sharing is considered an effective means of cost saving, it also runs the risk of propagating a particular failure across multiple products when the platform is defectively designed. Thus, sharing a common platform effectively can amplify the risk of incurring a large number of failures. In this paper, we formulate a quantitative model for assessing amplified failure risk. Our analysis shows an unexpected result that the platform risk increases as we become more assured of our design capability; i.e., when a defective design rarely happens, the magnitude of risk amplification becomes much larger. The numerical investigation of a platform planning case sheds light on the importance of risk assessment in determining the level of commonality in designing a platform.Index Terms-Decisions under risk and uncertainty, decision analysis, mathematical tools for management of technology, new product development process, R&D management.

Sourcing assemble-to-order inventories under supplier risk uncertainty

“…Yim [51] proposed a multiple-supplier procurementallocation model that minimizes a risk-averse cost function that quadratically increases with the number of defects. Where a design error is responsible for defects, Kang et al [23] showed that the defect risk is exponentially amplified where a common design is shared among different products. It also suggests riskaverse manufacturers to diversify suppliers in order to adopt their independent designs.…”

“…In Kang et al [23], the defect probability of a component is assumed uncertain according to random occurrence of deficiency in its design. The number of defects is modeled as a mixture distribution of the design deficiency and defect events.…”

“…For manufacturers, product failure incurs significant quality costs in the forms of warranty repairs, recalls, or lost reputation [28], which can in fact be much larger than costs from unreceived orders. As Yim [51] and Kang et al [23] emphasized, supplier diversification is an effective strategy for mitigating this risk, since defects of components are correlated when they are supplied from the same supplier.…”

Vulnerability diffusions in software product networks

“…This study explores interplays between vulnerability diffusion, ND, GD, and SDR as observed in SPNs. Existing studies have applied conventional analytical modeling approaches to network analysis (Gokpinar et al, 2010; Kang et al, 2015; Sosa et al, 2013) that usually cannot incorporate SPN dynamics. In particular, the diffusion of innovations view (Mitra & Ransbotham, 2015) fails to consider the dynamic nature of software ecosystems, and the epidemic view neglects transitive relationships (Newman, 2018).…”