Header image

Case Study for Implementing AI/ML in Magnetic Particle Inspection

Tracks
NDT UNLOCKED
Knowledge Level - NDT Level III
Presentation Topic Level - Advanced
Target Audience - General Interest
Target Audience - Level III Managers
Target Audience- NDT Engineers
Wednesday, October 8, 2025
3:30 PM - 4:00 PM
Monterey 2-3

Speaker

Agenda Item Image
Mr Randall Smith
Vice President Of Quality and ESG
Jost International

Case Study for Implementing AI/ML in Magnetic Particle Inspection

Presentation Description

Human error can be removed from the dark booth magnetic particle visual inspection step and replaced by 100% AI/ML camera inspection with results having historic traceability. The AI/ML is being deployed for MT inspection taking place in a manufacturing setting by the Digital Transformation using cameras with lens filters designed to block UV light reflection. Teaching the AI/ML system requires an annotation/feedback loop using MT experts and computer screens capable of showing real time results from the system. After the AI/ML detector system has learned enough to start inspection, master parts with known defects are used as a means of calibration for the system. With the AI/ML system trained to correctly identify indications, the system has the ability for production integration and automation with results stored electronically for full digital traceability.



Short Course Description

Biography

Vice President of Quality for Jost International. Dedicated over 30 years to quality system management, supplier development and continuous improvement, to include over 10 years of experience overseeing Jost’s NDT department with ASNT level III certification in UT and MT. Recently focusing on reducing process risks and production internal PPM failure rates by leveraging AI/ML to perform inspections.
loading