A Decision Tool for the Supplier Selection of Automotive Back-End Semiconductor Intelligent Manufacturing

Abstract

The COVID-19 pandemic has resulted in a significant disruption of the semiconductor manufacturing supply chain in the automotive industry. The disruption is primarily due to the long cycle time and stringent quality requirements necessary to fulfill the zero-defect mindset. Also, the rapid growth of the autonomous vehicle market where Advanced Driver-Assistance Systems (ADAS) are crucial for enabling autonomous capability and relying on automotive optical sensors and back-end semiconductor manufacturing to fabricate these key components.

The selection of the supply chain is a vital component in guaranteeing that semiconductor components are delivered on time and that quality criteria are met simultaneously. Traditional selection criteria include quality, delivery, price, logistics support, and capacity. However, Industry 4.0 introduces new technologies to enable intelligent factory elements such as high-speed computing, internet connectivity, machine learning algorithms, and advanced AI applications. Consequently, supplier analysis incorporates an emerging digitalization factor to assess readiness for the new quality mindset, upgrade process automation, and establish advanced IT systems for product traceability and data analysis.

The Analytic Hierarchy Process (AHP) is successful for resolving problems with top-to-bottom hierarchies, assuming independence among layers. In contrast, Analytic Network Process (ANP) facilitates the development of more intricate, interdependent linkages and response including the elements of the hierarchy but requires more data and presents greater implementation challenges than AHP. The research verifies the compatibility and correction between results from the AHP and ANP models using Excel and Super Decisions tools, as well as consistency ratio (CR) checks to ensure effective implementation of the final decision-making tool.