This is 911 — Yes, we were expecting your call.
The Use of Artificial Intelligence to Predict Medical Emergencies
HLSensory Overload: We’re Everywhere You’re Going To Be
Like a scene out of Minority Report, computers are being used to predict human activity. But in this case rather than predicting a crime, they are predicting medical emergencies — with startling accuracy. A new and constantly improving artificial intelligence system is allowing ambulance companies to predict where the next call is going to come from, and will even suggest moving an ambulance in that direction before the caller even picks up the phone.
Emergency medical service (EMS) providers need to get to the scene of an emergency quickly and efficiently to help those in need of life-saving care or to help someone who is unable to care for themself. But how do they do it?
The days of relying on maps, EMS responder’s knowledge of the local area, and the use of lights & sirens are being replaced by surveillance technology that is now at the core of making you safer. Using global positioning systems, analyzing past 911 response call history, leveraging traffic cameras, road congestion, and data transmitted from cellular telephones — predictive modeling software can now let EMS managers deploy ambulances more efficiently based on knowing where the next 911 call will most likely occur.
Various consumer based software systems collect data from many separate sources that are then analyzed using artificial intelligence. The resulting analysis creates an “amoeba” that is placed on a map that indicates the likelihood of an emergency call occurring in a given area. Other displayed attributes include response time estimates, best route selection, and coverage recommendations.
The advantages of using predeictive moding system is that emergency management planners can more precisely deploy ambulances in areas that are most likely to experience requests for service. Targeted system deployment saves cost by reducing fuel consumption, limiting carbon emissions, limits unnecessary vehicle wear by moving from one standby location to another, and decreases crew fatigue by limiting the required time spent inside a cramped ambulance. Most importantly, those requiring emergency assistance receive services faster when units have be placed close to a potential call and are directed to the call using the fastest route at the time of the response resulting in the arrival of a professional rescuer quicker.
Predictive modeling solutions become “smarter” the longer that they run and as new and diverse data streams are connected. The speed by which this artificial intelligence can make recommendations far surpasses what the human brain could process. In addition, the more inputs that are analyzed, the more effective the system will become. For example, adding traffic flow, camera systems, weather, road work, and historical data will enable these systems to provide more accurate and targeted automated recommendations to the human ambulance operators.
What do you think? Is this a good use of data or does it infringe on personal privacy? Should it be a person’s right to opt out of predictive algorithms if opting out has a negative effect on everyone else in the system?
