(Image courtesy of NASA.)
When firefighters enter a building, they’re expected to use their training and senses to find trapped civilians and deliver them from danger. While drilled behaviors and instinct are important tools for every firefighter, they can’t compare to the insight that can be gleaned from big data.
Over the last 9 months the US Department of Homeland Security and NASA’s Jet Propulsion Laboratory have been hard at work developing an artificial intelligence system that can leverage big data to help keep firefighters safe.
Named the Assistant for Understanding Data through Reasoning, Extraction, and sYnthesis (AUDREY) this algorithm can track firefighters as they move through a structure using sensors embedded in the first-responders’ uniforms.
“As a firefighter moves through an environment, AUDREY could send alerts through a mobile device or head-mounted display," said Mark James of JPL, lead scientist for the AUDREY project.
Armed with a suite of sensors that can detect heat in adjacent rooms, concentrations of dangerous gases, and detailed maps of a structure, firefighters would be able to move through a structure in the safest, most efficient manner, making it possible to save more lives and protect their own.
But sensors alone aren’t enough to make AUDREY work. The brains of the AUDREY AI systems run on the cloud, leveraging computing power and the system’s ability to learn and make predictions about what first responders will need in the immediate future.
Though it’s only a few months old, the AUDREY system has already been tested in a virtual demonstration at the Public Safety Broadband Stakeholder Meeting held in San Diego. During the test AUDREY was given data from a number of different sensors and was expected to give recommendations to a group of phantom first responders via mobile device. While JPL didn’t explicitly say the test went well, Edward Chow, manager of JPL's Civil Program Office did say that within a year AUDREY will begin field demonstrations.