Big Data and the Future of Firefighting
How firefighters are putting big data to work to prevent fires and improve response times.
New York City Fire Department, United States
Nobody knows where or when the next fire is going to break out, but big data analytics is helping to predict risk and detect locations where a fire is most likely to occur.
New York City uses a regression analysis tool called FireCast that assesses more than 7,500 risk factors (such as a buildings construction materials, location and tax status) for the more than 300,000 buildings it inspects. The program predicts where a fire is likely to occur using information that comes from 17 agencies as well as from calls to municipal services.
Every day firefighters receive a list of the top 15 at-risk buildings in their district. According to fire chiefs, big data has improved the accuracy of its fire risk inspections by about 20%.
The fire department in New Orleans launched an outreach program in 2015 that leveraged analytics to prioritize those neighborhoods that were least likely to have smoke alarms and most likely to experience fire fatalities. Once identified free smoke alarms were distributed.
Big data is also being used to help predict and fight wildfires. For example, the Los Angeles Fire Department now uses a fire analysis system called WIFIRE to not only monitor a fire but predict where it will go and what it will do next. The web-based platform combines satellite imagery and real-time data from cameras and sensors to produce a picture of the fire, the conditions around it and its likely trajectory.
In 2017 during the Thomas Fire, one of the largest wildfires in recent Californian history firefighters used WIFIRE to monitor the blaze and prepare for what it might have done next. The system updated every 15 minutes. Meanwhile, the public logged onto the app to determine whether they needed to evacuate.
Data Rich Fire Rescue
The City of Amsterdam Fire Department in the Netherlands has been working with data for several years. Its first project was nicknamed 'The Firebrary' and involved building a standard library of firefighting terms which was shared between fire departments to ensure that everyone was speaking the same language.
The department is also using predictive modeling to build models for assessing risk. The data comes from numerous sources such as personal protective equipment and fire hose sensors. Additionally, new fire engines for the city are 'open data fire engines' that collect and send information such as how much water they're using and how much fuel is in their tanks.
Improving Response Times
Big data and data analysis should be able to help in other ways too including improving response times. Data can be collected and analyzed regarding factors that influence the speed of response such as traffic conditions, the state of roads and population density. If evidence is uncovered that hinders response times such as potholes civic authorities may be convinced to do something about them.
Next Story »