Data-Based Approach for Final Product Quality Inspection: Application to a Semiconductor Industry

The early information about the health state of the final product quality plays a vital role in the intact management of production. In semiconductor manufacturing, quality control of a too-small number of wafers is routinely carried on specific metrology stations, and the obtained quality measurements are generalized over the entire lot. The unavailability of sufficient product quality information results in a lack of that for a high proportion of products. The latter leads to some overlooked quality problems that might cause a malfunction in the final product. This malfunction is usually conducive to yield loss, resource consumption through its remaining production line steps and also needs a considerable amount of time to be source-identified. This paper proposes a final quality classification data-driven approach using machine learning techniques and alarm events data collected during the production operations. We use the k-mean clustering algorithm to group production lots into clusters based on their passages over equipment. Each cluster has its decision tree classification model elaborated after various information extraction techniques and manipulation applied to alarm event texts. The obtained results show a satisfactory performance demonstrated on a real-world dataset collected over the whole semiconductor fabrication facility.

I. Introduction

In the semiconductor manufacturing industry, Process control and final products’ quality are intimately linked during manufacturing, and their enhancement directly impacts productivity and quality. Therefore, semiconductor facility owners use the best possible solutions to have effective process control and monitoring systems that allow them to achieve productivity targets, final product quality requirements, desired production costs, and delivery times. The alarm system constitutes an essential part of the facility automation system and a significant role in manufacturing operations monitoring. It communicates to the control system and the operator any abnormal situation, process deviation, or equipment malfunctions that requires a timely response [1]. According to a predefined configuration and limits, these abnormal situations or faults are determined using equipment sensor real-time data. They might also be considered as sources of quality problems that are usually detected at the final test or after a customer return. The final quality prediction based on collected data history is very considerable to provide early information about the final product health state and support decision making.