ABSTRACT
The aim of this chapter is to provide an overview of automated optical inspection (AOI) edge artificial intelligence (AI) inference system solutions in the digital industry by considering if, and how, they enable manufacturers to reach a satisfactory trade-off between customer needs and production costs. Numerous solutions can address customer and factory needs, from inspection machines to testing boards equipped with cameras installed near the conveyor belt. In all the considered solutions we can implement effective defect detection algorithms, such as the latest You Only Look Once (YOLO) variants based on deep learning (DL), to obtain high key performance indicators (KPIs), i.e., mean average precision, adequate process capability and high throughput yield. Parallel implementations of edge test systems allow us to further improve production yield, while repeated tests performed in sequence can allow us to approach the precision required for zero defect practice. The comparison of available solutions using KPIs, functional requirements (FRs) and non-functional requirements (NFRs) highlights that the advantage of using inspection machines is that they are equipped with user interface and data analysis which helps workers and managers to ensure high quality production process and effective order management. Their weakness is the computing boards for defect testing at the edge are featured by lower costs. A demonstrator to evaluate the effectiveness of edge AI solutions based on the test boards available on the market and those developed by the EdgeAI project is outlined.
