How C2SMARTER helped FDNY understand post-COVID emergency patterns with AirSage 

To understand what was happening when the emergency response vehicle crossed an intersection, Jannie and her team used a match process. They tried to match the FDNY data and AirSage data based on road segment at a certain 30 second time aggregation. They were able to match about 1,000 events between FDNY and AirSage data during a two-week period: 500 matches while the emergency response vehicle was enroute to the accident site, and 500 matches as it was going to the hospital with the patient.

Jannie’s team then looked at vehicle speed during the 30 seconds when the data matched to see what was happening. They focused on life-threatening incidents only for this comparison since the emergency response vehicle had very different behavior than during non-life-threatening incidents.

Because the AirSage data and FDNY data did not report speed in the same way, Jannie’s team did some post-processing of the AirSage data to ensure the speed calculations were an apples-to-apples comparison. They wanted to understand if there was a difference between the two parties. Their hypothesis was that most likely the general public had lower speed, and the emergency response vehicle had higher speed.

For the matches, the difference in the speeds between the two sets of data ranged from about 8 – 17 miles per hour (Figure 1) depending on whether the trip was enroute to the incident site or was going to the hospital. Thus, by looking at the AirSage data, they could prove that emergency response vehicles were going much faster than civilian vehicles.