CloudFAS: Cloud-based building fire alarm system using Building Information Modelling

Abstract

Building fires are a common urban disaster. The emergence of high-rise, large-scale and inner-complex buildings bring new challenges for fire safety and triggers new demand to upgrade traditional building fire alarm system (FAS). Different from current studies by deploying enormous smart fire sensors to replace FAS, this study addresses this issue from a novel perspective and proposes a cloud-based FAS using Building Information Modelling (BIM) on top of FAS, termed CloudFAS. Firstly, the system framework and the software architecture are designed. Secondly, two key technologies are presented to address two unresolved technical issues: private fire alarm data sharing and alignment of fire sensors with the BIM model. A cloud gateway for fire sensors is developed to address the first problem by capturing the fire alarm data from the fire alarm control unit through the IEEE 1824 standard. Noticing that the fire sensor locations are listed in a sensor installation spreadsheet using natural language, termed as sensor location table (SLT). A natural language processing (NLP)-based sensor-BIM alignment algorithm is proposed to automatically match fire sensors with the BIM model through SLT, which enables to display fire sensor statuses in proper places in the 3D BIM model. Finally, a concrete case study from the China Construction Library is presented, which verifies the effectiveness of our proposed CloudFAS. Our CloudFAS is built on top of traditional FAS. If the fire alarm control unit follows the IEEE 1824 standard and an SLT is available, then CloudFAS can upgrade the traditional FAS in existing buildings effortlessly with its BIM model. Moreover, the cloud gateway for fire sensors contributes to addressing the private data sharing problem using IEEE 1824 standard, and the NLP-based sensor-BIM alignment algorithm can promote the adoption of BIM in the building operation phase.