The main innovation is the shift from reactive, breakdown-based maintenance to predictive management that identifies warning signs before failure. Rather than relying only on routine inspections, the project continuously collects vibration and noise data, analyses equipment condition and remaining life, and connects the results to remote monitoring and field maintenance-history management. It is distinctive because diagnosis, history management, parts-cycle analysis and technical-document control are integrated in one platform.

The project is also innovative in scale and institutionalisation. It moved beyond a pilot at one site, expanded from Line 1 to Line 2, and is now being rolled out on Line 7. By combining AIoT, big data and mobile maintenance functions in an operating railway environment, it offers a practical and expandable public-sector model for smart infrastructure maintenance.

The system has been completed for Lines 1 and 2 and is now being expanded to Line 7. By 2024, it had been installed on 190 HVAC units on Lines 1 and 2. In 2025-2026, procurement, contracting and field installation are under way for 28 units across six stations on the Bucheon section of Line 7. Expansion to additional sections is planned, so implementation and diffusion are progressing together.

Incheon Transit Corporation defined the operational problem and led the project, while private technology firms supplied sensors, gateways, analytics and system integration. KC Mirae Technology Co., Ltd., identified as a joint patent holder, acted as a key technology partner. Agreements with the Korean Society for Artificial Intelligence Education and AI Works also supported talent development and training, strengthening technical reliability, field applicability and future diffusion.

Direct users are maintenance staff, depots, mechanical teams and control personnel responsible for HVAC inspection and repair. They use the platform for remote checks, pre-inspection review and maintenance-history analysis. Stakeholders include the corporation's departments, partner companies and academic institutions supporting system development, training and diffusion. Final beneficiaries are subway passengers, who benefit from more comfortable stations and more reliable HVAC service.

The innovation aims to improve HVAC reliability, manage underground-station air quality more systematically, and make maintenance work more efficient. Deployment has been completed on Lines 1 and 2 and is expanding to Line 7. By 2024, 190 HVAC units on Lines 1 and 2 had been equipped, and 28 additional units are being added on Line 7 in 2025-2026. The project has also produced two patents, five academic outputs and a 2024 Top 10 Railway Technologies award. Impact is evidenced by deployment scale, expansion, research and award results, and documented gains in work efficiency and station comfort.

A major challenge was embedding a new technology in a live railway environment while expanding it line by line. The preparation phase required a testbed and performance verification, and the expansion phase faced differences in operating conditions and budget availability across lines. In particular, the Incheon section of Line 7 had not yet secured a budget, limiting immediate rollout. To manage this, the corporation used a phased strategy-preparation, Line 1, Line 2 and Line 7-reducing risk through verification, integrated monitoring and staged expansion.

Several conditions supported success. First, the project aligned closely with national policy directions such as the Basic Plan for Mechanical Equipment Development and measures to improve underground-station air quality, giving it a strong public rationale. Second, a phased approach-from testbed to verified expansion-built technical trust. Third, an integrated AIoT infrastructure combining sensors, communications, analytics, monitoring and mobile history management made field use practical. Fourth, collaboration among the public sector, companies and academia linked technology development, implementation and training.

First, public-facility maintenance reaches clear limits when it depends mainly on routine inspections and experience; data-driven predictive management improves effectiveness and sustainability. Second, new technology is better introduced through a testbed, performance verification and phased expansion than through one-time full deployment. Third, predictive maintenance is not achieved by sensors and algorithms alone; it must also include history management, parts-cycle analysis, manuals and training. Fourth, collaboration among public agencies, companies and academia strengthens both technical outcomes and diffusion potential.

This innovation is more than a facility-management system; it is a scalable model for the digital transformation of railway maintenance. Two patents, five academic outputs and recognition among the 2024 Top 10 Railway Technologies show both technical maturity and external credibility. The establishment of a permanent training site and talent-development partnerships also indicates that the system is being embedded as an organisational capability rather than remaining a one-off project. It has strong potential for transfer to other metro operators and similar public infrastructures.