Kuken Industries searches 70 years of drawings in minutes

Kuken Industries, a major Japanese manufacturer of cooling towers and ventilation equipment, faced challenges in locating replacement part drawings from its 70-year archive. Engineers, especially newer ones, relied on experienced colleagues to find and interpret these drawings, which could take half a day. To address this, Kuken partnered with AI startup Recursive and OpenAI to develop an AI-powered search system that allows engineers to search by part number or upload images/PDFs. The tool extracts details and returns possible matches, reportedly reducing search time by about 90%. This enables newer engineers to work more independently, frees experienced engineers for mentoring and engineering work, and supports knowledge transfer. Kuken is also prototyping an AI tool to review drawings for discrepancies, with about 90% of flagged issues being genuine, aiming to further reduce review time and preserve engineering expertise for future generations.

(Source：OpenAI)