Envision being trapped underneath the wreckage of a hurricane with no phone lines. What if social media could save your life by giving you an online platform with a more accurate and detailed route of the hurricane?
Purdue University researchers are now testing technology to help first responders find victims of natural and man-made disasters. An online platform called the Social Media Analytics and Reporting Toolkit (SMART) could help first responders monitor areas where hurricanes are more likely to hit, and allow them to monitor social media posts to find people in need of aid.
“SMART distills the ocean of social media data down to relevant and usable information in real time,” said David Ebert, director of Purdue’s Visual Analytics for Command, Control and Interoperability Environments and the Silicon Valley Professor of Electrical and Computer Engineering. “What an incredible opportunity to use our engineering talents at Purdue to make a life-saving impact.”
The system has not yet been released publicly, but official first-responder organizations can request an account to use SMART during Hurricane Florence by contacting Ebert at email@example.com.
First Responders can also use the platform to select key words such as “flooding” or “medical emergencies,” which are then highlighted on a map as they are talked about on social media in a specific geographic area.
“Imagine being trapped after a hurricane makes landfall and having social media as the only form of communication to reach out for help,” Ebert said. “SMART allows first responders to pick up those digital cries for help and take action.”
The technology also allows users to set up customizable email alerts for relevant key words within a specific time frame. Officials in California used SMART to find victims from this year’s wildfires.
SMART can also be used during potential attacks during major public events such as analyzing school threats or monitoring traffic.
“There is really no learning curve because the interface is visual and natural,” Ebert said.