University of Waterloo researchers’ new framework could improve evacuations, inform building codes and help emergency responders better understand toxic smoke hazards
A team of researchers is harnessing artificial intelligence (AI), data science and advanced mathematics to better predict how fires behave in modern homes, work that could ultimately help save lives during emergencies.
The new system, developed at the University of Waterloo, addresses a growing challenge for fire researchers: modern furniture materials and increasingly airtight, energy-efficient homes can cause fires to behave differently than they did in the past.
“The goal is to deepen our understanding so we can predict fire behaviour and the gases it releases, enabling smart systems that support safer and more effective fire evacuations,” said Dr. Joshua Pulsipher, a chemical engineering professor at Waterloo.
To generate the data needed for the project, researchers conducted 15 experimental fires in a burn house on campus, using up to 175 sensors to measure temperature, airflow, humidity, burn rate and numerous gases.
Even a single set of sensors in one location, sampling four times per second, generated millions of data points, representing only a small fraction of the data collected across all experiments.
Modern furniture and energy-efficient architecture were the focus of the study because new foams and fabrics produce different toxic gases, while oxygen-starved fires in airtight residences compound the danger by increasing the volume and changing the composition of those gases.
To make sense of the enormously complex information provided by sensors, the research team created a system that combines math, data science and machine-learning AI. The resulting analysis yields insights with the potential to inform building codes, better prepare emergency responders and plan evacuation routes.
“We want to create smart systems that model and anticipate what a fire will do and then route people to get out of the building safely,” said Dr. Beth Weckman, professor of mechanical and mechatronics engineering.
The system uses statistical relationships in the data to determine, for example, when a fire begins to under‑ventilate, or run out of oxygen, which marks a crucial change in combustion chemistry linked to the production of more toxic smoke.
“The input is the data from the experiments,” said Dr. Vinny Gupta, a mechanical and mechatronics engineering professor and a member with Weckman of the Fire Research Group at Waterloo. “The output is understanding what those experiments tell us to reveal fundamental insights about how the underlying fire evolves.”
In future research, the researchers plan to extend their system to analyze more complex fire scenarios.
Their study, A framework for high-dimensional fire sensor data analysis, was recently published in the Fire Safety Journal.
About the University of Waterloo
University of Waterloo is a leading global innovation hub that drives economic and social prosperity for Canada and the world. With more than 41,000 students, we are home to the world’s largest co-op education talent pipeline, to game-changing research and technology, and to an unmatched entrepreneurial culture. Together, these create partnerships and solutions to tackle today’s and tomorrow’s challenges. Find out more at uwaterloo.ca
