A statistical prediction model for the inspection process in an outbound automotive supply chain.

摘要

Commensurate with other disciplines, "supply chain management" is not oblivious to the "mantra" of continuous improvement. Consequently, all supply chain characteristics are constantly monitored to identify inherent process inefficiencies and unearth opportunities for improvement.;Automotive supply chains can be broadly classified into two categories, namely inbound process and outbound process. This thesis focuses on outbound logistics, specifically in the realm of vehicle inspections. The manufacturing process and the transportation of the finished vehicles can cause damages to the exterior of the vehicle necessitating inspection at the plant and at each supply chain node.;There are two functionally disparate albeit inter-related deliverables from this thesis. The first is a prediction model to prognosticate the areas of the vehicle prone to defects at the plant. The second part of the thesis employs an advanced statistical technique called "Capture-Recapture Modeling" to estimate the dealer claims based on the inspection data.