More than a thousand United Hatzalah volunteers were drafted into the IDF reserves as doctors, medics and in other critical roles after Hamas-led terrorists invaded the northwestern Negev on Oct. 7, 2023.
Nearly 14% of the emergency medical service’s manpower was gone, hurting response times.
Dov Maisel, the VP of operations, says the challenge forced him to look into ways to bridge the gap, ultimately leading him to leverage technology.
Veterans of the IDF’s technical intelligence Unit 8200 and members of United Hatzalah’s Operations and Technology Department were able to create an Artificial Intelligence system using 20 years of data based on the 2,000 emergency calls made every day. During a three-month pilot program, the system achieved 85% accuracy in predicting locations and times of emergency events, the organization says.
In Rishon Letzion
A month ago in Rishon Letzion, Maisel had a few hours of down time and decided to take a shift on an ambulance.
After checking in with dispatch, which is using brand-new AI technology to predict where emergencies might take place, Maisel and his team were advised to station the ambulance on a street near Azrieli Rishonim Mall in case an emergency happened. After less than five minutes of driving and arriving at the mall, the team was notified that an emergency was taking place near them.
“I don’t even put the ambulance into the parking space yet, and the app goes off and 400 meters away from me, someone is having chest pain,” says Maisel. “Within one minute, we pull up and the family is still on the phone with dispatch.”
Within seconds, the team was able to ascertain that the victim was suffering from a heart attack and begin treating him with advanced life support and medication.
“It’s simply phenomenal how we’re not only able to reduce response time, but use our resources in a smarter way,” says Maisel.
The AI predictive modeling used by United Hatzalah is designed to dramatically reduce response times by pre-positioning crews in locations where they often see emergencies, based on location and time of day.
The nonprofit is known for having a quick response time because of its coalition of volunteers stationed around Israel. Shifting model from reactive to proactive, the AI technology uses a heat map to show hot spots in city-specific locations, which dispatch and volunteers can access from their United Hatzalah application on their cellphones.
Within a few months of launching the technology in the United Hatzalah system, they’re already seeing a high success rate, Maisel says.
“We’re at about 90-95% accuracy in predicting where the next call will happen,” he says. “We’re getting endless accounts from the volunteers themselves.”
Maisel said the typical model of volunteers responding to emergencies isn’t changing, but the new technology will reduce driving and response time while improving safety for emergency crews.
Hundreds of United Hatzalah volunteers responded to the Oct. 7 massacre in southern Israel, with many of them saving lives under fire from Gazan terrorists. For the last year and a half, the organization has learned from that day.
“October 7 pushed us as an organization to limits we never imagined we would be in,” says Maisel. “Growing out of that, looking at the growing development of the war in the north, we understood that if we don’t prepare for everything on our own, nobody will prepare us. If the worst ever happens again, we’re ready for it.”
United Hatzalah trained more than 1,000 volunteers during the past year, and has more than 8,000 volunteers nationwide.