TaiwanManufacturing2022 – ongoingCheng Shin Rubber
AI inspection of tires for air-bubble defects
Cheng Shin Rubber moved tire bubble inspection from three shifts to one with Azure AI vision
Problem
Tires were checked for air-bubble defects by inspection staff working three shifts.
Approach
Cheng Shin worked with Microsoft and 兆米智慧檢測 at its Taiwan plant. Tire images go to an on-site Azure Stack Edge device, where an Azure AI Custom Vision model decides whether a tire has a bubble defect. Results go back to the line to remove defective tires and staff double-check. When new tire lines or formulas arrive, images are uploaded to Azure to retrain the model, which is pushed back to the edge.
Results
Takeaway
Keeping inference on a local device kept latency low on the line while retraining ran in the cloud. The source is an article sponsored by Microsoft Taiwan.
Categories
- Size
- Large enterprise
- Industry
- Manufacturing
- Business function
- Production and quality
- Innovation type
- Process innovation
- AI technology
- Computer vision
- Adoption
- Vendor solution
- Evidence
- Vendor case study
Sources
Source summaries
Sponsored article on Cheng Shin's AI bubble detection built with Microsoft and 兆米智慧檢測. It describes the edge and cloud setup using Azure Stack Edge and Custom Vision. It gives accuracy above 99.9% and a staffing cut from three shifts to one. Plans for overseas plants are not counted as results.
Full details are in the original sources.