I like to think of artificial intelligence (AI) as an Excel spreadsheet come to life: predefined variables based on computer science and data to mimic the human thought processes via the use of pattern recognition derived from statistics as they relate to the topic at hand. This then can be categorized as predictive (combining historical data to create algorithms that can “predict”) or generative (using language, audio, images or synthetic data to automate and accelerate). In EMS providers’ world, the latter can be used to allocate resources, thereby streamlining a better workforce.
AI can be viewed as a component of EMS that allows us to control what we can control—for example, the best road route to take to a call. The effect on response time that results from road closures because of construction or flooding and from time of year (e.g., surges that can occur during the summer and amid Christmas shopping season) can be mitigated when AI is applied.
AI also can be employed to improve EMS response/performance in conjunction with coordination with other agencies (the need for electric power to be shut off or availability of a warming bus that can function as triage or a holding area at a large-scale incident) and for the purpose of helping callers to prepare for the arrival of the ambulance via predictabiliyt of human behavior and emotional response.
Furthermore, AI can serve as support for responders, so they can focus on their expertise based on the formulation of best practices that often are developed during the “after-action” component of an incident.
Already, paramedics use computerized guidance in real time to track vital signs and compare patient history with current symptoms, which can assist in treatment of patients. In the future, computerized guidance—likely augmented somehow with AI—could lead to the development of the autonomous ambulance (in other words, no driver needed), which would allow two paramedics to remain with a patient to deliver better care.
There isn’t any reason to believe that AI won’t help EMS in the vein of the analysis of wait times, equipment failure and vehicle out-of-service time, and “nuisance calls” (e.g., the “chest pain” patient).
The balance
Despite all that’s noted above, what we know in EMS is that many of the problems that must be solved can’t be entered into a computer to establish data to formulate a response: Predictable human behavior in the prehospital world is most unpredictable.
Can technology alone save lives? Inherently, human interaction harmonizes technological advancements. It’s difficult for AI to make a split-second decision or to provide the physical ability to work to save a person’s life.
Think of it this way: Yes, the dishwasher at the firehouse can have sensors that contribute to optimal cleaning that’s based on whether just one rack or the whole machine is filled. However, the dishes must be loaded into the dishwasher and unloaded.
