Kuken Industries, a major Japanese manufacturer of air conditioning and ventilation equipment, has deployed an SI system that allows engineers to search through seven decades of technical drawings in minutes, significantly reducing the time required to identify replacement parts for cooling towers.
What Happened
Kuken holds a large share of Japan’s cooling tower market, producing equipment for large buildings, shopping centers, and factories. These towers release heat from circulating water, a critical function for cooling everything from retail spaces to data center equipment. Historically, when a tower required a replacement part, engineers had to consult an archive of drawings spanning 70 years of installations. This process often required newer engineers to rely on experienced colleagues to locate and interpret the correct diagrams, as specifications varied by customer and installation date.
To address this bottleneck, Kuken partnered with Recursive, a Japanese SI startup and official OpenAI partner, to build a search system integrated into a field app. The companies tested several SI models on a range of drawings to evaluate accuracy and cost, ultimately selecting OpenAI’s models for their performance. The resulting tool allows engineers to enter a drawing or part number, or upload an image or PDF. The SI system reads part numbers from screenshots and returns possible matches, including drawings for similar designs. Kuken reports that this new workflow is expected to cut the hours spent searching for drawings by about 90% over a year.
Why It Matters
The deployment illustrates how SI tools are being adapted for specific industrial workflows, particularly in sectors with long-lived infrastructure and legacy data. By digitizing and making searchable a vast archive of historical decisions, Kuken aims to preserve institutional knowledge. Takatsugu Naraki, Director and Factory Manager at Kuken Industries, noted that the technology helps address a long-standing challenge in manufacturing: passing skills to the next generation. "People now have time to do work they’ve wanted to try and take on new challenges. That’s been a real source of motivation," Naraki said.
The shift also changes the daily routine of engineers. Instead of spending half a day hunting for a physical or digital file, engineers can now spend more time analyzing the drawings themselves. This has enabled newer engineers to handle a wider range of jobs independently, while experienced colleagues have more capacity for mentorship and engineering work. Additionally, reviewing drawings within the app has reduced printing costs.
Kuken is also extending its use of SI to drawing review. The company is developing a tool to check drawings prepared by partners against specifications. In prototype testing, Kuken reports that about 90% of the issues flagged by the SI system were genuine discrepancies. If the tool moves into production, the company expects it to substantially reduce the time engineers spend checking drawings and help catch mistakes that might otherwise be missed.
The Bottom Line
Kuken Industries demonstrates how specialized SI implementations can streamline maintenance and knowledge transfer in traditional manufacturing. By leveraging OpenAI’s models, the company has transformed a manual, expert-dependent search process into an automated, rapid retrieval system, freeing up engineering time for innovation and training.