TaiwanManufacturing2020 – ongoingFIH Mobile
AutoML Vision defect inspection for smartphone circuit boards
FIH Mobile cut the share of missed circuit-board defects from 40% to 10% with machine-learning visual inspection
Problem
Humans inspected printed circuit boards before production, and 40% of defective boards slipped through uncaught. FIH Mobile, a Foxconn group handset maker, wanted a faster and more consistent check.
Approach
FIH Mobile built an automated system on Google Cloud's AutoML Vision and finished the deployment in April 2020. An image capture rig with two microscopes and an X-Y table moves each component set through inspection, and the models classify each component as pass or fail. All inspection stations use the same process.
Results
Takeaway
The team only had to prepare and label the images, and the same models gave every station the same standard. Missed defects fell by three quarters.
Categories
- Size
- Large enterprise
- Industry
- Manufacturing
- Business function
- Production and quality
- Innovation type
- Process innovation
- AI technology
- Computer vision
- Adoption
- Off-the-shelf tool or platform
- Evidence
- Company's own claim
Sources
Source summaries
Google Cloud's customer story describes how FIH Mobile used AutoML Vision to inspect smartphone circuit-board components for defects. A company manager reported the defect escape rate fell from 40% to 10% and inspection took 0.3 seconds per component. The page carries no date; the Internet Archive first saved it on July 9, 2021.
Full details are in the original sources.