Deep learning based multi-platform report models for smart factory

Abstract

Highly advanced modern Information Technologies are causing big changes in mass production industries and opening $4^{th}$ Industrial Revolution (Industry 4.0) which are based on intelligence smart factory. Artificial Intelligence is expected to do her duties not only on production automation and management, but also production data interpretation and inference which are traditionally considered as the rights that adhere to man. In this thesis, the author proposes the machine learning based methods which execute characterized products data classification and deduction in semiconductor production line. And also suggest modeling which could increase production yield and maintenance efficiency on this basis. Additionally, we study the translated data conversion to post final report types which are suitable for human recognition.