Conference Notice

Optimization of fireground operations is a key challenge for fire agencies, particularly as wildfires trend towards more destructive. As response resources become stretched, trade-offs in safety and effectiveness will require resources to do more with less. Empirical data on fireground operations is important for decision support in these cases, but challenging to collect in active emergency situations. In Victoria, Australia, we have been utilizing data science techniques to transform operational data sources into tools and evidence to support our fire agencies in decision-making. In this talk, I will demonstrate some of our novel approaches, such as using Automatic Vehicle Location data from fire trucks and aircraft to quantify fire control production rates and time spent on activities. We have also used fire reporting data to develop machine learning models to assist with triaging of fires when there are multiple wildfire ignitions. This program of work has been very valuable to help fire managers to better understand their current operations, and how we can improve in the future.