Digital Twin's AI and VR boost Efficiency in Building Energy and Maintenance

Digital Twin's AI and VR boost Efficiency in Building Energy and Maintenance

When it comes to the design and construction of green buildings, we will definitely think of the emerging architectural concept of Building Information Modeling (BIM). However, green buildings emphasise sustainable management methods and energy utilisation. Among them, there is a technology that is rarely mentioned in the market but complements BIM, which is Digital Twin.

What is Digital Twin Technology?

Professor Xiao (middle) and two doctoral students who are members of her research team, Chen Zhe (right) and Ma Tianyou (left).

Digital Twin Technology refers to reproducing the objects that exist in reality for the construction of an extremely detailed and highly similar virtual model, and then establishing a digital identity for each corresponding object. The two exist separately like "Twins" in real and virtual worlds. The main function of Digital Twin is to collect real-time data in the real world for machine learning and data analysis, in order to simulate and predict the possible reactions and implications of objects in the real world, as well as studying their effectiveness. This involves the Internet of Things (IoT) technology used to collect data, multi-source heterogeneous data fusion technology, and even the concept of unmanned management.

Digital Twin was first applied in industries such as aerospace. Nevertheless, as technologies such as artificial intelligence (AI) and Virtual Reality/Augmented Reality/Mixed Reality (VR/AR/MR) have become increasingly mature and popular, Digital Twin have gradually become an important technology for managing green buildings and the corresponding energy systems.

In this regard, Professor Xiao Fu, Associate Dean of the Faculty of Construction and Environment at The Hong Kong Polytechnic University (PolyU), led a team to develop a Digital Twin system and applied it to the energy optimisation, and operation and maintenance of the centralised air-conditioning system in the PolyU campus. By collecting real-time operational data and using AI technology to analyse data, the system has been combined with physical and engineering knowledge to provide optimised operational strategies and predictive maintenance schedules that aim to enhance energy efficiency. Integrating MR, BIM, Building Automation System (BAS), IoT and multi-source heterogeneous data fusion, this Digital Twin system has been developed for two platforms: Windows and Hololens 2.

Professor Xiao’s team is developing and testing the digital twin platform for smart building management.

Digital Twin Technology and its Application Value

"After wearing mixed reality glasses, the air-conditioning system installed in the wall or ceiling will become visible to the naked eye. With the developed indoor navigation technology, maintenance personnel can be guided to the malfunctional equipment from any location within the building. Upon arrival, they can see the surroundings, while using a virtual interface to look into the operating parameters of equipment from sensors, as well as the results of AI-empowered data analysis. This makes it easier for them to make real-time decisions.” This smart and convenient system is designed to be user-friendly, even for maintenance personnel without any AI backgrounds.

To accurately monitor the real-time conditions of each equipment, it is crucial to precisely map the real-time operational data from a large number of sensors into virtual models. The digital objects within these models need to perfectly match the corresponding physical equipment in both temporal and spatial dimensions in the real world. The Digital Twin Technology developed by Professor Xiao's team excellently addresses this challenge.

 

"The digital twin system we have developed is an advanced smart building energy management system designed for existing BAS. Through wired or wireless communication, the system collects real-time data from various equipment in the building, including the equipment at the chiller plant and the indoor environment. The data is visualised in a 3D interface based on BIM, facilitating the remote monitoring of equipment operations. By leveraging AI technologies, particularly image recognition and machine learning methods integrated with physical and engineering knowledge, the system performs data analysis and sends optimised control signals to BAS to minimise energy consumption while ensuring occupant comfort. Professor Xiao continued, "For the chiller plant, we have developed AI-empowered strategies for chiller sequencing strategy, chilled water temperature setpoint reset strategy, and predictive maintenance strategies for chillers and pumps. Meanwhile, for the space air conditioning system, we have developed AI-based strategies for indoor parameter setpoint reset strategies and predictive maintenance strategies for supply air terminals. These strategies help to improve the overall energy efficiency of the building energy systems and reduce carbon emissions, while ensuring thermal comfort and air quality ".

User interface of the digital twin platform for smart building management (Windows version).

BIM model of the chiller plant in PolyU campus (Phase 1/2/2A/2B).

Visualisation of equipment real-time data.

Real-time AI recommendation for chiller on-off control.

Application of mixed reality technology for on-site operation and maintenance work.

In addition, building occupants can submit their personal perceptions of and preferences for indoor thermal comfort to the digital twin system by scanning a QR code. By utilising AI algorithms that integrate physical and engineering knowledge to recommend the setpoints of indoor temperature and reset the fresh air volume, the system enables human-in-the-loop control so as to enhance occupants’ experiences. "To address privacy concerns, the data collected via the IoT currently does not include any user identification information. Submitting feedback through QR codes assists us in providing a more personalised indoor environment and optimising energy use," Professor Xiao stated.

Development Direction of Digital Twin Technology

Currently, in addition to the PolyU campus, this digital twin system has been partially applied to a government office building in Hong Kong and the Midea Headquarters Building in Guangdong. Students in Professor Xiao’s team have achieved impressive results in a number of innovation and entrepreneurship competitions and have been invited to participate in several exhibitions.

The team was awarded the 1st Runner-up Award at the Hong Kong Techathon 2023.

The team won the First Prize in The First Greater Bay Area Data Application Innovation Competition.

"We have validated the feasibility of the technology through various application cases, accumulating significant experience and data that can help to improve the energy efficiency of building energy systems, and facilitate the operation and maintenance of building electrical and mechanical systems. With the integration of Internet and cloud computing technologies, the service location of the digital twin system is no longer limited. Our plan is to initially promote its application in commercial and office buildings," Professor Xiao added.