General Science launches AI model for tire-quality control

Yinchuan, China – Chinese tire group General Science has unveiled a new AI-based quality-control system which it described as a “milestone leap” for digitalised tire manufacturing.

Developed with Tsinghua University’s Vehicle & Mobility School and Yinchuan-based Beirui Technology, “Huiyan AI large model” was launched in mid-November, said General Science 21 Nov.

The project aims to “break industry bottlenecks” in "tire-appearance inspection", a process long dependent on manual checks and prone to “personnel fluctuations, inconsistent standards and accuracy limits.”

Describing itself "an early mover in smart manufacturing", the tire maker said it has already introduced a semi-steel fully-automated “dark factory” - an operation that does not need human involvement - and a full-steel “5G carbon-cloud smart plant.”

The AI platform is intended to further advance its model of “smart manufacturing + green production.”

The new system integrates AI and augmented-reality tools with “leading optical-electrical structures” and Tsinghua-backed theory, enabling full data sharing and interlinked quality monitoring.

At the launch, the companies presented the first ‘LingShi AI tire inspection terminal’ based on the Huiyan model.

The terminal uses a “large-sample self-iterating algorithm,” enabling “out-of-the-box” use without additional training data.

It also allows “seamless switching” between tire types, shortening deployment times.

According to the presentation, inspection time per tire has been reduced to 25 seconds, compared with an industry norm of about 30 seconds.

System-wide detection efficiency is 194% higher than conventional setups while missed-defect rates are reported at 2.7%, and below 1% for functional defects.

The terminal also links real-time quality signals across six upstream processes, from mixing to curing, creating a full loop of “self-perception, deep analysis, interactive feedback and iterative improvement.”

General Science general manager Gu Cui said the model is the first milestone of the company's smart-manufacturing strategy and offers a template that is “replicable and scalable.”

The partners have also signed a strategic cooperation agreement to expand the use of the AI model in industrial scales.