Advancing Fire Safety: A Digital Twin-Based Framework for Smart Buildings

By Samudrapom DamReviewed by Susha Cheriyedath, M.Sc.Dec 19 2023

In an article recently published in the journal Buildings, researchers proposed a digital twin (DT)-based fire safety management (FSM) framework for smart buildings that leverages several state-of-the-art technologies, including artificial intelligence (AI), to improve FSM.

Background

In recent years, the rising trend of computerization and digitization has enabled the application of information technologies in facility management (FM) to enhance fire emergency practices. However, multiple challenges, such as the inefficiency of conventional methods utilized for information retrieval in fire safety equipment (FSE), must be addressed to ensure optimal FSM in buildings.

Thus, preparing real-time information-based intelligent FSM is crucial to overcome these challenges by improving the decision-making process and providing access to information for safe fire evacuation and FSE maintenance in real time. DT technologies with real-time data integration and acquisition can be developed to realize this goal.

Although several industries have implemented DT technology in the last few years, the FSM sector has been slow in adopting this technology. Key DT-enabling technologies for FSM include augmented reality (AR), AI, the Internet of Things (IoT), and building information modeling (BIM).

The proposed FSM framework

In this study, researchers proposed a DT-based FSM framework for smart buildings by leveraging key state-of-the-art DT-enabling technologies for FSM. The intelligent framework can offer a high degree of adaptability and flexibility by serving two separate applications focusing on FSE maintenance and fire evacuation practices. The DT building FSM involved several layers, including a fire warning control layer, a smart platform FSE maintenance and fire evacuation layer, a DT data layer from BIM and IoT data, a monitoring layer, and a physical layer.

Researchers structured the proposed framework into four layers, including the user interaction layer, the application layer, the virtual building layer, and the physical building layer. The framework operated through a centralized physical building layer, which acted as the key source for real-time data obtained from IoT devices, indoor localization systems, and fire sensors.

Then, these data were processed and analyzed in the virtual building layer using predictive algorithms for fire spread and evacuation simulations. In this virtual layer, data-level fusion integrated real-time sensor data with predictive algorithms for decision-making, while decision-level fusion occurred in the application layer where several functionalities, such as occupant tracking, evacuation path planning, and fire detection, were synthesized for optimizing emergency responses..

Eventually, the user interaction layer utilized the AR technology to provide users, such as occupants, firefighters, and facility managers, an engaging visualization of the building and its safety features. The virtual building layer integrated with the application layer seamlessly by leveraging AR, enabling users to make informed modeling decisions presented in a visually interactive format in the user interaction layer.

Data were collected using sensing technologies and transmitted to the virtual building layer through a wireless network in the physical building layer, while the IoT, BIM, and FSE data were processed using AI tools in the virtual building layer. The FSE maintenance modeling in the virtual layer used AI techniques, such as deep learning (DL) and machine learning (ML), to perform data analysis to optimize and classify the FSE maintenance work order. Additionally, data fusion techniques were utilized to integrate data from different sources, and blockchain was employed to ensure data privacy and security.

All decisions made in the virtual layer were integrated back with the physical building layer to realize automatic control and operation optimization and transferred to the application layer to provide services. The objective of the application layer was to integrate, analyze, store, and visualize data models and platforms using AI technologies, simulations, and ML.

Smart technologies like AR were used to visualize decisions from the application layer to integrate end users. All decisions were confirmed and checked by FM professional users before their implementation to ensure human involvement in the process.

Evaluation and findings

A questionnaire was conducted for FM professionals to comprehensively evaluate the proposed DT framework, specifically the framework’s data security and clarity for fire evacuation planning and FSE maintenance. The survey results demonstrated that the proposed DT framework can effectively assist decision-makers in obtaining comprehensive information about facilities’ communication among stakeholders.

Almost 76% of the respondents strongly agreed or agreed that the proposed framework clearly described the potential use of fire evacuation planning and the FSE of facilities/FSE maintenance, while 24% of the respondents remained neutral. Similarly, 58% of the respondents strongly agreed or agreed that the data workflow in the proposed DT framework was secure, while 33% remained neutral. Only 9% of the respondents strongly disagreed or disagreed with the statement. These results indicated that the proposed DT framework was secure and effective in the context of FSE maintenance and facility fire evacuations.