Digitizing 70 Years of Drawings: How a Cooling Equipment Maker Applies AI to Skills Transmission

Kuken Kogyō, a manufacturer of cooling and air-conditioning equipment with a 70-year history, addresses the challenge of passing experienced engineers' skills and know-how to younger staff by combining its archived drawings with AI. In collaboration with Recursive, an OpenAI partner, the company is building a system that makes its vast paper records searchable for anyone.

■ The Skills Transmission Challenge in Manufacturing
- Why inherited expertise is not reaching younger employees smoothly
- Cooling towers used in large public spaces such as museums and shopping centers
- Structure differences between older and current equipment, and how experience guides document selection

■ Building the Drawing Search System
- Digitalizing 70 years of paper documents and the limits of simple digitization
- Searching archive documents by part numbers entered by users
- Selecting OpenAI models based on internal benchmark results for unstructured data

■ Operational Impact and Human-in-the-Loop
- Shortened drawing search time and how the freed-up time is allocated
- Keeping final decisions with people while AI supports thinking time

Manufacturing, DX, and AI implementation practitioners can gain a concrete view of how AI is applied to legacy unstructured records while preserving human judgment. Consider reviewing your own organization's knowledge archives and identifying where similar search support could be applied.