When disaster strikes, 911 dispatchers may quickly be inundated with phone calls, and people in desperate situations may face long waits before someone answers and dispatches emergency responders.
Ademola Adesokan, a Ph.D. student in computer science at Missouri S&T who is set to graduate in December, says he feels so passionate about this issue that he dedicated his doctoral research to a method to support the 911 system.
“When people call 911, they are usually in a perilous situation — often between life and death — and need help quickly,” he says. “But depending on the quantity of calls coming in at once and the amount of time spent on each call at the 911 center, they may have a delayed response during a time in which every second or minute can seem like an eternity.”
Adesokan says most people use smart phones to place 911 calls, and they often already have social media applications downloaded.
“Based off this notion, I developed and trained a model that can track emergency posts people make on social media and then route the information to the appropriate agencies,” he says. “People have already used social media during emergencies for years, but there has not been an interface to quickly fetch, process and perform different tasks before sending this information where it needs to go.”
Adesokan says he trained the model with social media posts from recent natural disasters, as well as the COVID-19 pandemic. The system he developed can review the content of the posts, including their text, emojis, hashtags, and any other relevant information. It analyzes the emotions in the posts and can identify more detailed information about the situation that first responders will find beneficial.
When multiple entities in one area use the system, it would automatically identify pertinent information about the situation, assign the information a classification and send it to whichever outfit was identified as the most appropriate for the specific issue.
“Maybe someone is trapped in their home after a flood, and their phone only has a few minutes left before the battery dies,” he says. “They may not be able to stay on the phone long enough to reach a 911 dispatcher, but they or someone else who knows their whereabouts could post on a social media platform, and whichever agency is primarily handling rescues would receive an alert.”
This model could also be used in the event of a school shooting or during another time in which people may need to quietly share their situation and cannot place a phone call.
Adesokan says agencies using this system would see an interface on their devices and not have to be on the actual social media pages. However, they could see the original posts and potentially respond to them, depending on the situation.
“This wouldn’t be a replacement for the current 911 system,” Adesokan says. “However, it could significantly support it and get the information more quickly to the proper people.”
He says a challenge with this type of setup is the amount of irrelevant information, or noise, people may post about an issue — or even “fake news” that may be shared. However, his model is designed to cut down on those issues and identify characteristics of posts that may fit into these categories.
“These issues already exist, and first responders, when they check social media, often have to filter through large amounts of posts that slow down their ability to be helpful,” he says. “Although we cannot control all of the noise on social media, my model can at least work to identify what information is most relevant and urgent and then give it the right classification to go to the right people.”
Adesokan says the technology he developed holds a special place in his heart because it was inspired by someone close to him passing away when he was nine years old and living in Nigeria, and he is hopeful it will one day be adopted by agencies on a large scale.
“Someone very dear to me had a stroke, and the emergency response time was delayed and very slow,” he says. “With that type of medical event, people need care immediately to have the best chances of recovery. I want other people to have opportunities to have faster care that were unfortunately not available for my loved one.”
Adesokan, who will be S&T’s first graduate of the Kummer Innovation and Entrepreneurship Doctoral Fellows Program, says he could not have conducted his research without the support he received as an I&E Fellow, along with the support of his advisor, Dr. Sanjay Madria, a Curators’ Distinguished Professor of computer science. The support also led to him winning the National Science Foundation’s Spirit of I-Corps Award for his research this summer.
“As an I&E Fellow, I had the freedom to research a topic that was so important to me,” he says. “The funding and support students receive through this program is truly life-changing and helped shape my future for the better.”
