There are numerous kinds of E&M systems in buildings, which generate massive amounts of operational data, making manual maintenance both strenuous and inefficient. To this end, the EMSD developed “Engentica”, an AI agent for E&M systems in buildings, significantly enhancing the efficiency of daily maintenance. In November this year, the project won the Best Use of Technology in the Built Environment and the Digital Champion Award at the Society of Digital Engineering Digital Awards 2025 organised by the Chartered Institution of Building Services Engineers (CIBSE), making the EMSD the first Hong Kong government department to receive the champion award.
To effectively manage the wide range of E&M equipment in buildings, the EMSD has over the years been actively promoting smart management, from the early establishment of a one-stop integrated Building Management System to the setup of the Regional Digital Control Centre in recent years for remote monitoring and data analysis. With the rise of generative AI models such as DeepSeek, the EMSD also began to develop an AI platform, “Engentica”, which was designed specifically for E&M systems and officially launched in March this year. “Engentica” connects to the EMSD’s database, covering real-time operational data, maintenance records, work schedules and various operation manuals. The system can quickly integrate and analyse data to assist in executing multiple tasks, such as fault diagnosis, scheduling, material review and energy optimisation, revolutionising building management practices.
Using chiller plant replacement as an example, engineers traditionally had to analyse large volumes of operational data and maintenance records manually. They would then apply various calculation tools to estimate plant performance, degradation, and the payback period for replacement costs, making it a complex and time-consuming process. With “Engentica”, engineers can simply enter the relevant commands and the system automatically performs calculations and generates reports, significantly enhancing work efficiency. When a fault occurs in an E&M system, “Engentica” quickly retrieves operational data and inspection procedures from its database, analyses the root cause of the fault, and accelerates repair progress.
“Engentica” has been piloted in five government venues, including office buildings, laboratories, and data centres. Since its launch, it has increased fault-resolution speed by 40%, reduced administrative workload by 30%, and achieved around 10% energy savings.
We also plan to collaborate with a private data centre and a shopping mall to further test and optimise the system. The team will continue to enhance “Engentica”’s hardware configuration and system functionalities to support the management of more government venues. In the long term, we hope to share this AI platform with the industry to promote smart E&M services.
