Bushfire building risk assessment using advanced technologies

Overview

Bushfires often have detrimental impacts on both the natural and built environments. Although current building standards are in place to reduce the influence bushfire has on new buildings, existing and older residential properties are more susceptible to bushfire ignition. Identifying and restoring the most vulnerable features in existing properties can assist in reducing the property damage caused by bushfires. Implementing new technology into these risk assessments of existing bushfire-prone properties can reduce the time required and lower costs. This study aims to develop a model to assess the risk of individual existing residential properties by identifying their surroundings using satellite imagery and machine learning. It uses geospatial data to determine the vegetation density surrounding houses in bushfire-prone areas. A sample of satellite images from two suburbs (ie. near Brisbane) will be annotated into three categories – trees, grass, and non-vegetation. These results contribute to developing a machine-learning model that minimises the time spent identifying vegetation density surrounding individual building structures. This model, combined with heat transfer models, will allow for cost-effective bushfire building risk assessments for individual existing residential properties in bushfire-prone locations and reduce damage.

Civil/Mechanical/Electrical Engineering students interested in advanced technology, data analyses and/or technical writing are required. The potential student will work with the QUT Wind and Fire Lab researchers.

Research engagement

• Review the process of pre-bushfire condition assessment methods
• Use satellite imagery to identify the vegetation density surrounding houses
• Reconstruct the buildings (ie. houses) and conduct heat transfer analysis for the selected buildings in Brisbane, Queensland
• Investigate the use of satellite data for better bushfire planning and prediction

Outcomes

• The study will contribute to developing guidelines for using already available advanced technology in pre-bushfire condition assessments of buildings.

Skills and experience

Civil/Mechanical/Electrical Engineering students interested in advanced technology, data analyses and/or technical writing are required. The potential student will work with the QUT Wind and Fire Lab researchers.

Start date

25 November, 2024

End date

7 February, 2025

Location

GP - Campus

Keywords

• Bushfire
• Buildings
• Fire resistance
• Wildfires

Contact

Dr Anthony Ariyanayagam - Email: a.ariyanayagam@qut.edu.au Tel : 07 3138 251

Prof Mahen Mahendran - Email: m.mahendran@qut.edu.au Tel : 07 3138 2543