Artificial Intelligence tool for fire safety design (IFETool): Demonstration in large open spaces

Abstract

Fire modelling is a common practice in building fire safety analysis, but it is costly. This work develops an AI software, Intelligent Fire Engineering Tool (IFETool), to speed up the fire safety analysis and quickly identify design limits. A big numerical atrium-fire database is firstly formed by considering key building and fire parameters. Then, a deep learning model is trained to predict the evolution of tenable smoke visibility, temperature and CO concentration with an accuracy of 97%. The tenability descending profile is further processed to assess the available safe egress time (ASET) and the fire safety of the atriums that have complex roof shapes and slab extensions. This AI design software is able to make a quick assessment of the proposed atrium fire engineering design and give valuable suggestions for potential improvement. Finally, the operation guidelines of IFETool are provided for common design tasks of atrium fire safety.

Keywords

• Smart building
• Fire engineering
• Intelligent design
• Deep learning
• AI Software
• Atrium