Building Semantic Artificial Intelligence (REF:C-0037)
E&M InnoPortal Trial Project Ref. No.:

Overview

Hong Kong has been actively developing into a smart city in recent years. The Electrical and Mechanical Services Department (EMSD), responsible for providing electrical and mechanical engineering services for over 8,000 government buildings, plays a crucial role in implementing Hong Kong's smart city blueprint and the digitisation of electrical and mechanical systems.

However, developing artificial intelligence applications for building systems has always been a complex task as it often involves a lot of components and systems. One of the major problems is the non-standardised naming convention and classification for the building's electrical and mechanical (E&M) equipment. This lack of standardisation makes it difficult to integrate data sources and apply knowledge gained from one building to the others. Furthermore, undesirable repetitive manual data processing and model training are required which would largely increase the time and cost required for technology adoption.

The "Building Semantic Artificial Intelligence" solution was developed by EMSD to address the challenge of deploying artificial intelligence models across different buildings along with their E&M systems. The solution allows standardisation of data model for various E&M systems, such as air conditioning system, through a common semantic artificial intelligence model. By combining technologies of digitisation and natural language processing (NLP), the solution thereby enables porting artificial intelligence model among various buildings.

The semantic AI model allows for analysing the seamless analysis and application of operating data from one building to others. As such, it significantly reduces the time for developing customised AI applications for each building by 70%, from 1-1.5 year down to 3 months. As a result, it optimises system operations, saves energy, and establishes a solid foundation for the digitalisation of urban management.

Problem Addressed

In this solution, it aims to introduce a common semantic AI model to represent building systems in a consistent and standardized format. By aligning semantic representations, the solution facilitates widespread data-driven optimisation of building operations and maintenance through artificial intelligence.
Innovation

This project pioneered the innovative application of building service deployment by orchestrating ontology-based semantic modeling, machine learning and knowledge graph techniques to revolutionise and scale up the automated deployment of AI-powered solutions across building portfolios in Hong Kong.

By developing the first standardised framework to represent building equipment systems, domain experts can now extract and integrate asset data programmatically and automatically via a unified semantic model platform.

This novel project is transformative as it is the first-time ever that these technologies are integrated and creatively applied to automate AI-enabled applications among large numbers of buildings. As the pioneering work in this field, this semantic AI approach transforms how AI techniques can be massively deployed to enhance energy use, maintenance operations and asset management for sustainable facilities management citywide.
Key Benefits
  • Enables mass deployment of AI across a large number of buildings with improved efficiencies, collaborative capabilities and data management, thereby largely reduces the deployment time and cost
  • Enhances collaboration between stakeholders by sharing of knowledge base via the semantic data platform with unified asset representations
  • Provides comprehensive digital asset management supported by real-time data, as well as building operational insights through semantic AI referencing to showcase the development of technological creativity in Hong Kong
Patent and Award
  • International Exhibition of Inventions of Geneva in 2021 (Gold Medal)

Building Semantic Artificial Intelligence: The Future of Automation in City Level - a real application in today

  • The 48th International Exhibition of Inventions of Geneva (Bronze Medal)

City-level Robotic Inspection & Diagnostic System for Building Systems
Project Reference West Kowloon Government Offices, Tai Lung Veterinary Laboratory
Other Reference
  • Global AI Challenge for Building E&M Facilities

https://globalaichallenge.com/en/home 

  • Video Introduction

https://youtu.be/q-mUO99kGrw