Deep-i, a non-destructive testing startup, has signed a business agreement with automotive aluminium casting parts manufacturer Hanju Light Metal to develop and deploy an AI-powered X-ray automatic inspection system.
Under the agreement, Deep-i will install its AI-based automatic defect-detection system at Hanju Light Metal’s production sites. The partnership covers the full process, including system development, on-site deployment, demonstration and performance verification.
The collaboration is focused on automating the inspection of automotive aluminium casting components, where X-ray testing is used to identify internal defects such as pores and cracks.
AI-based system to automate X-ray defect detection
The system being developed by Deep-i consists of three main components: an AI model for automatically detecting and judging defects from X-ray inspection data, an integrated control platform for real-time inspection-result monitoring and historical data management, and an AIOps system that enables Hanju Light Metal to upgrade the AI model independently.
The system is designed to address the limitations of the existing inspection approach, which relies on skilled workers to visually interpret X-ray images. By applying AI-based analysis, the companies aim to establish consistent defect-judgement standards while enabling systematic management of inspection records.
For automotive aluminium casting parts, X-ray inspection plays an important role in identifying internal discontinuities that may not be visible through conventional visual inspection. The inspection data can reveal defects including pores and cracks within cast components.
Deep-i has developed solutions for automating inspection processes in manufacturing environments based on its technology for analysing non-destructive testing data and automatically detecting defects.
Partnership targets automated NDT inspection
The agreement will see the two companies work together across system development and implementation, as well as demonstration and performance verification at Hanju Light Metal’s production sites.
The planned AI-powered inspection system is intended to combine automated X-ray image analysis with real-time inspection-result management and AI model management, creating an integrated approach to automated NDT data analysis.
"We will build an AI automatic defect-detection system that actually works on the ground and create a case of cooperation in which both companies grow together," said Kim Ki-soo, CEO of Deep-i.
The partnership represents a move toward automated X-ray inspection for automotive aluminium casting production, with AI being applied to the analysis and classification of NDT inspection data.
Reference: https://en.sedaily.com/technology/2026/08/14/deep-i-hanju-light-metal-team-up-for-ai-x-ray-defect
