← AI Inspection Atlas

Automotive · Production · Government news (Chinese)

Autoliv (Jiangsu): in-house AI vision system inspects laser welds on burst discs for automotive safety components

Company / plantAutoliv (Jiangsu) Automotive Safety Components Co., Ltd., Jintan District, Changzhou (China)
Defect typeLaser weld defects (varied, complex weld defect shapes)
PartBurst discs (rupture discs) for automotive safety components
MaterialNot stated in source
VendorAutoliv (in-house algorithm model)
Camera / lightingNot stated in source
Training dataAlgorithm model built by the team and tuned repeatedly with industrial shop-floor data
Reported result (quoted)No inspection metrics published. The system was selected for Changzhou NDRC's "2026 second batch AI+ application scenario list." Deployment uses a centralized server architecture in which one server handles detection inference for several production lines at once. The source says it makes up for manual inspection's unstable recognition of tiny flaws.
NotesJintan District government site (district media center), Nov 16, 2026, in Chinese. The source notes conventional machine vision suits rule-based features such as size, position and count but struggles with variable weld defects; the AI system targets that gap and runs on the production floor. Autoliv plans further AI uses in automotive safety components.
Defect categoriesWeld & spatter
Sourcehttps://www.jintan.gov.cn/html/czjt/2026/IJLEQALK_1116/304793.html

Results are quoted from the source. Vendor-published results are vendor claims and were not independently verified. Spot an error? Contact us.