AI is Poised to Bring Improvements in Public Safety & Emergency Rescues
It’s no secret that, according to VcloudNews, humans and machines generate 2.5 quintillion bytes of information each day on the Internet. How big is this? It would fill 100 million Blu-ray discs, which if stacked, would measure the height of four Eiffel Towers on top of each other. With this much data being generated, we’ve reached a point in our lives when humans can no longer be expected to process information at a usable rate by themselves. This is where the “Age of Artificial Intelligence” needs to come into existence if we are going to utilize information in any useful way.
Understandably, AI scares people. We are often afraid of what we don’t understand. However, when used properly, AI can be an assistive technology and not in control. Intelligence is already built into some of the most common events in our lives. For example, after washing, I put my clothes in the dryer. I use the automatic setting, and press start. I’m now free to walk away, go for dinner, or even take a nap if I choose. The dryer is “programmed” to run for a time, and on a newer machine, a sensor indicates the moisture content of my delicate items. In either case, when criteria are met, the heat shuts off and my clothes continue to tumble. This simple logic has made my dumb clothes dryer “artificially intelligent.”
AI can be applied to any process, including many predefined procedures found in public safety 911 centers. Remember though, this is not about replacing the call taker or dispatcher—it is about providing them with relevant additional information, based on predetermined indicators, helping them arrive at a decision quickly. Let’s look at two outcomes of the same scenario, first using today’s technology, and then using artificial intelligence to augment and assist in the decision-making process.
Car vs Deer
David is speeding in his 2018 GM vehicle. Susan is sitting next to him in the passenger seat. A deer runs out hitting him head-on. David swerves into the median, the car overturns four times and lands on its roof leaving both occupants unconscious. Sensors in the vehicle detect a high Delta-V (rate of velocity deceleration), both passenger and driver airbags deployed, and specific crush zones on the vehicle.
The in-vehicle system (IVS) generates a call to the OnStar call center, flagged as an emergency. The call is routed to an ACD queue staffed by emergency medical dispatchers. Vehicle data and location information appears on their screen. An attempt to communicate with the occupants verbally is initiated, and a three-way conference with the public safety agency responsible is started. Information is passed on verbally to the 911 call center, where local protocol for dispatch is followed.
The same situation initiates a different process or workflow. In addition to notifying OnStar and attempting to get a call taker in verbal contact with the vehicle, a SIP session is set up in the Emergency Services IP Network (ESInet) where the 911 call taker is presented with the telematics data and bridged into a three way-audio bridge with the vehicle and the OnStar call taker. The system analyzes the data and determines an 80% chance of entrapment and lower leg trauma.
The dispatcher is prompted to dispatch the recommended resources, which include heavy rescue, advanced life support, and a medical air unit. They also can edit resources desired and dispatch with a single button.
Bed counts and staffing levels are examined at local hospitals, the availability of an orthopedic surgeon and operating room is determined, and based on big data, a destination facility is recommended. A single touch to confirm or edit, and the data is on its way to the hospital where staff can prepare for patient arrival.
This situation has brought to light the efficient use of AI to determine the best response and action, all while remaining under complete control of a human. Resources become more efficient and effective and are available sooner for other missions. While many may be afraid of AI replacing humans, thanks to Arnold Schwarzenegger in the Terminator movies, I don’t see Skynet being right around the corner.
An added benefit is that AI is available from the cloud, making it affordable to agencies of all sizes. It’s not about building a data center to process data in the building anymore. It’s about using the cloud through multiple resilient paths, sharing the workload with other agencies that will also be available to provide coverage and backup for when “the big one” hits, no matter where or when that might be. This radically changes the curriculum for a Public Safety career, however the skill sets required are also taught for positions in the commercial space, and best practices remain across verticals.
As I’ve said before, AI is not just about HAL, and getting pod bay doors open. (Besides, in addition to being intelligent, HAL copped an attitude, but in reality, he was just programmed that way.)