Sign-up deadline: August 31, 2023

 

Call-for-Participation Webinar

Download presentation (July 11 & 14, 2023)
 

Statement of Work and Project Statement

• Statement of Work (V2.1, June 13, 2023)
• Project Statement (V2.1, June 13, 2023)

 

Background

Machine learning and artificial intelligence (AI) models have been adopted to help with the pass/fail judgement of automated optical inspection (AOI) machines. Large quantities of images are used to train AI models to identify defects. Integrating AOI data to appropriate predictive analytic solutions and incorporating AI solutions into AOI processes has the potential to detect true failures with greater accuracy and in less time. It also helps reduce man-hours, re-inspection and repeated machine program optimization processes.
 
The iNEMI AI Enhancement to AOI for PCBA Project aims to establish common language and performance metrics to evaluate the value of adding AI to AOI capabilities and provide recommendations to the industry on future implementation and improvement opportunities.
 
In Phase 1, an industry-wide survey was conducted to assess the adoption status and challenges of AI for AOI in board assembly. Based on survey results, the project team designed a test vehicle incorporating the most interesting defect and component types and ran an experimental trial on AI training and AOI inspection.
 
The project is now calling for industry participation in Phase 2 experiments, which will use the established test vehicles for training and/or inspection using existing AI for AOI solutions. Collaboratively, this project will help industry mitigate technology gaps and improve inspection efficiency and accuracy as well as reduce cost.
 

Contact

Haley Fu
haley.fu@inemi.org