JapanManufacturing2024 – ongoingNSK Ltd.
In-house generative AI application that visualizes and summarizes past quality-trouble data
NSK gave over 5,000 domestic employees a generative AI app that summarizes about 4,000 quality records in about 30 seconds
Problem
NSK's past quality-trouble data sat in dedicated databases and reports and was hard to read without knowing product-specific terms.
Approach
NSK began technical validation of generative AI on table-structured quality data in October 2024. Using agile development, it moved from functional design to official deployment in about six months. The application shows graphs of accumulated quality records and writes short AI summaries. It runs in NSK's own environment under AI quality-control measures and guidelines for risks such as hallucinations.
Results
Takeaway
NSK reported access and response time but no savings or quality results. It planned to extend the tool to sales and logistics and to overseas sites.
Categories
- Size
- Large enterprise
- Industry
- Manufacturing
- Business function
- Production and qualityInternal work and knowledge management
- Innovation type
- Process innovation
- AI technology
- Generative AI
- Adoption
- Built in-house
- Evidence
- Company's own claim
Sources
Source summaries
NSK's release on its first full-scale internal generative AI application. It gives the record count, user group, response time and development period, and describes the risk controls around the tool.
Full details are in the original sources.