KT Cloud Unveils Data Center Operations and Construction Innovations Using AI and BIM

KT Cloud shared innovative cases applying AI and BIM, based on 25 years of accumulated data center operation experience. Managing approximately 56,000 devices and 2.98 million monitoring points across nationwide data centers, it generates over 9 trillion data points annually, leveraging this to shift operations from Reactive to Proactive.

Operator-Led AI Tool Development Platform

Through AIDC Ops Studio, where AI generates code when on-site operators describe functions in natural language, five tools have been developed for proactive response to anomalies, predictive maintenance, and cooling energy efficiency analysis. In particular, it comprehensively analyzes server usage patterns and weather forecasts during cooling equipment operation to determine optimal operating levels and simulate energy savings. Operators can directly create tools and deploy them to other centers; utilizing FastAPI, Next.js, and Google Gemini series models, simple automation can be deployed in one day, while complex analysis tools take 1–2 weeks.

BIM-Based Construction and Digital Twin Expansion

In the design and construction phases, BIM (Building Information Modeling) is used to pre-review interferences between facilities and reduce rework. By combining LiDAR 3D scanning and AR technology to verify construction results, and integrating BIM data with real-time operational data, a strategy to expand into Digital Twin is being pursued. This minimizes design errors and establishes a data-driven decision-making system that continuously utilizes asset information even after completion.