Part II: Innovations
February 21, 2020
Charleston County Consolidated 9-1-1 Center
Editor's Note: This is the second installment in a five-part series describing innovations in emergency communications.
An imperfect set of circumstances turned into an incredible opportunity to solve a problem that can be applied universally.
The circumstances are common: Centers are understaffed yet responsible for handling all incoming calls, including the disproportional number of alarm calls. There are no easy solutions, particularly considering a projected 6% increase in dispatch job growth over the next decade fueled by rapid turnover1, and an increasing rate of false alarms from citizen and business security systems.
The imbalance wasn’t going away anytime soon.
“That’s what drove us,” said James Lake, director, Charleston County Consolidated 9-1-1 Center in South Carolina (USA). “We looked at the workload and what we could do to modify what we did without affecting our service to the public.”
The problem wasn’t only the waste of department money and officers’ time. Law enforcement in Charleston County has a tracking policy allowing emergency dispatchers to ask officers to keep a running check at a building after a flagrant number of false alarms have been received from the same location. In addition, Charleston County charges a $50 fine for fourth, fifth, and sixth instances of false alarms and a $100 fine for seventh and all subsequent false alarms.
Keeping check and fines for excessive false alarms hardly made a dent in the number of false alarms coming into the center, which, said Lake, compromises 14% of all incoming calls. On average, only 1% are valid (real) alarms.
They still had to be answered and that’s where Lake realized that a solution relied in the how they were answered. With his deep technical understanding from 40 years in public safety, Lake looked at a combination of emerging technologies to create a new and efficient customer experience without sacrificing 911 service.
“If it can be done in the business world, why can’t it be done in the 911 world?” he said.
Lake ended up enlisting the aid of IBM Watson, a computing system that provides a precise answer to questions asked using natural language processing (NLP) algorithms. NLP builds on a machine’s ability to process what was said, structure the information received, determine the necessary response, and respond in a language that is understood.
Charleston County Consolidated 9-1-1 Center’s goal is to use IBM Watson – Voice to process alarm calls.
During the project’s initial stage, Watson was fed with hours and hours of alarm calls and the sequence of questions in the Police Priority Dispatch System™ (PPDS®) Protocol 104: Alarms Key Questions—such as type of alarm and name and address of business/resident/owner. When an alarm goes off, companies monitoring the security systems call a separate number set up at Charleston County Consolidated 9-1-1 and answer Watson’s questions scripted from protocol. The answers provided by the alarm company operator would be converted to text and used to create a CAD incident.
An escape mechanism, or bypass, allows Watson to transfer specific alarm types (i.e., panic and medical) to a live calltaker. During initial testing, simulated calls were monitored closely to check Watson’s answers. Analysis proved Watson was mostly right on target.
“There are a few bugs we to still iron out,” Lake said.
A refined Watson will be put to full use following a pilot project testing live calls.
The emergency dispatchers were hesitant. On the one hand, they were skeptical about trusting their job functions to an application existing in a cloud. On the other hand, they realized advantages to relinquishing a portion of their workload to Watson.
“They do an outstanding job, and Watson frees them up to help people who need them,” Lake said.
Lake doesn’t anticipate limiting Watson to alarm calls. For example, mobile devices equipped with artificial intelligence in police cars can vocalize text data and create dynamic incident reports. Because police car cameras are hardwired into a vehicle’s electrical systems, they’re able to activate based on triggers that are programmed into them. For example, if a police officer turns on lights and siren, a dash cam could automatically start recording the situation. Dispatch can run the license plate from the images provided to alert police officers of possible scene safety dangers even before the officer gets out of the squad car.
It’s hard to predict exactly where artificial intelligence will take Charleston County Consolidated 9-1-1, Lake said.
“Until we start looking and testing the options, we don’t know where it will go,” he said.
In other innovative news, Charleston County Consolidated 9-1-1 rolled out a website that’s the first of its kind. It manages high volume calls, handles situations that could potentially jeopardize the caller and bystanders, and acts as a substitute for limited or downed phone service.
Called 911HelpMe.com, the site routes emergency traffic to administrative telecommunicators—and not trained and certified emergency telecommunicators—via a short form completed and submitted online. Administrative telecommunicators use the information to begin processing the request, which includes choices of when to contact the individual.
The online system has several benefits, Lake said. It picks up on requests that could be missed during call volume spikes, protects people in compromised situations, and routes non-emergency calls made to the appropriate source for assistance.
1“Police, Fire, and Ambulance Dispatchers.” U.S. Bureau of Labor Statistics. Occupational Outlook Handbook. 2019; Sept. 4. https://www.bls.gov/ooh/office-and-administrative-support/police-fire-and-ambulance-dispatchers.htm#tab-1 (accessed Jan. 10, 2020).