Case Study: Hitachi achieves 100% defect detection in visual inspection with Google Cloud Platform

Hitachi Using visual inspection AI for quality control and achieving 100% accuracy

Hitachi, the Japanese electronics manufacturer behind Omika Works, needed a faster and more reliable way to inspect crimp terminals on control panels. Manual visual checks were time-consuming, prone to human error, and difficult to scale across factory operations. Hitachi turned to Google Cloud Platform and its Visual Inspection AI service, connected to the Hitachi industrial edge computer CE50-10, to simplify defect detection and make the system easier for on-site staff to use.

Google Cloud Platform helped Hitachi build a machine learning model from about 100 sample images in roughly one day, enabling automated inspection of scratches and assembly quality over a VPN-connected edge setup. The proof of concept achieved a 100% defect identification rate, with no false negatives, and was deemed usable for factory deployment. Hitachi also reported a 13% false reporting rate during testing, which it expects to improve further as the solution is tuned for real-world production.

Hitachi

Masayuki Takagaki

MONOZUKURI Management Department, Production Management Division