The Importance of Keeping Score

The Importance of Keeping Score

Unverified alarm notifications are plaguing law-enforcement and other emergency-response agencies nationwide—but a new approach to analyzing them offers a potential solution.

By Frank Fernandez, President of Blueprints 4 Safety (B4S) Strategies Group, and John Chiaramonte, ENP, Consulting Division President of Mission Critical Partners.

For more than a century, since the earliest pioneers in alarm monitoring began connecting their customers to public safety agencies, the task of separating false alarms from real emergencies has been a persistent challenge. On one hand, thousands of lives and millions of dollars in property are saved every year because of public safety’s response to monitored alarms, ranging from structure fires and carbon monoxide incidents to home/business invasions and burglaries/robberies. On the other hand, most alarms are unverified, and many turn out to be false. That creates a serious conundrum for the public safety community—respond or not?

To answer this vital question, the public safety community traditionally has erred on the side of caution by responding to unverified alarms. The theory is that it is far better to respond and not need to save a life than to not respond and lose one. However, public safety agencies today receive more 9-1-1 calls and operate with fewer staff than at nearly any time in recent history. Continuing to support communities with high rates of alarm notifications that may turn out to be false alarms is becoming more problematic. Consequently, some agencies have stopped responding to unverified alarms.

Most agencies, however, still respond to alarms without any additional details or verification. This is distressing not only from the perspective of responder effectiveness, but also safety and operational efficiency. When officers, firefighters, and emergency medical technicians (EMTs) are dispatched because of a false alarm, they are placed at unnecessary risk. For example, vehicles responding to an incident occasionally are involved in accidents; in some cases, first responders (or even civilians) have been seriously injured or killed. This exposes the agency and the municipality it serves to liability-related lawsuits totaling in the millions of dollars.

In addition, responders have been conditioned to think, based on experience, that when they are dispatched to an alarm call, that no “real” incident is occurring (i.e., a false call). The result is that a level of complacency sets in—which is exactly the wrong mindset to have while en route to a potential emergency. It is imperative that our responders are prepared and focused upon arrival.

Finally, when responders are sent to incidents that prove to be false, our limited resources are squandered in the sense that they are unavailable to respond to other legitimate events. This delays emergency response, which is a bad outcome when lives are on the line and every second matters, and when they do, they place personnel at unnecessary risk when the call turns out to be a false alarm. Worse, responders are unavailable for legitimate emergencies, which often increases response times.

The Partnership for Priority Verified Alarm Response (PPVAR) is working on a solution. PPVAR’s mission is to educate electronic security industry and public safety stakeholders on the merits of leveraging video and audio analytics to reduce false alarms and improve response outcomes, including the following:

• More efficient emergency-response dispatch
• Enhanced responder safety
• Better use of emergency-response resources
• More arrests

PPVAR’s idea is to work collaboratively with the alarm industry to reduce the number of unverified alarms—in particular false alarms—and to provide better situational-awareness information to 9-1-1 telecommunicators so that they can make well-informed response decisions, and to law-enforcement officers and fire/rescue/emergency medical personnel so that they have a better idea of what they’re going to face when they arrive at the incident scene.

To accomplish this, a framework for algorithms that will be used to score alarm notifications were developed by a PPVAR working group and shared with The Monitoring Association (TMA)—a trade group dedicated to the advancement of the professional-monitoring industry—for ratification as an industry standard. The alarm companies would apply the standard to analyze the data generated by their systems—supplemented by video and audio analytics—to determine the score, and then to contact the appropriate emergency-response agency if that is warranted. If so, the score will help the agency to triage the incident and determine what, if any, response is necessary.

That contact ideally would be accomplished via the ASAP-to-PSAP1 standard, which is a joint effort between TMA and the Association of Public-Safety Communications Officials (APCO). The standard enables machine-to-machine transactions between alarm companies and PSAPs, thereby reducing the number of voice calls that tie up phone lines and require interaction with a 9-1-1 telecommunicator. The ASAP-to-PSAP electronic notification automatically would enter the emergency-response agency’s computer-aided dispatch system. This is significant because today, every alarm notification generates from three to five 9-1-1 calls from the alarm company (between the original activation with multiple follow-up calls).

The PPVAR concept is similar to that developed by the telematics industry for vehicle crash-notification systems (like GM OnStar, Ford Sync and others). Such systems gather data generated by a variety of vehicle sensors, which then is weighed using an algorithm especially created for this purpose. The result is a score that predicts the severity of the crash and the likelihood of life-threatening injuries.

Here’s an example of how this can work. Let’s say that a vehicle crash occurs on a rural roadway, and the scene is 60 miles from the nearest fire/rescue/EMS agency. It will take an hour for emergency responders to arrive, and then some time for EMTs to assess the situation. Let’s also say that one of the victims has suffered life-threatening injuries. It will be another hour before the victim arrives at the nearest trauma center.

Now let’s consider the aforementioned crash-notification system, which has determined, based on the sensor data gathered and analyzed, that there is a very high likelihood that one victim has life-threatening injuries. Based on that score, a decision is immediately made to dispatch a medical-evacuation helicopter, which arrives on scene within 10 minutes. Fifteen minutes later, the victim arrives at the trauma center. This is a significant time reduction—less than 30 minutes versus two hours—that could be the difference between the victim surviving or not. Consider that the first hour after a traumatic injury has occurred—the so-called Golden Hour—is the most critical in terms of successful emergency treatment.

The scoring system envisioned by PPVAR will have similar impact for both the law-enforcement and fire/rescue/emergency medical communities. The alarm industry will analyze the data generated by a myriad of sensors, using the algorithms that are being developed, to determine the likelihood that any given alarm notification is an actual emergency and, more importantly, indicate the type and severity of the incident via the scoring mechanism. Here are a few examples of how this might work:

• When a bank robbery alarm is triggered, video captured by the bank’s video system could be analyzed to determine whether a robbery is in progress and—using a video-analytics-based algorithm driven by artificial intelligence—to determine the number of people inside the bank, the likelihood of the robbers being armed, and potentially even the type of weapons they have. Again, the availability of such situational-awareness information would have a significant bearing on the type of response dispatched and would prepare officers better for mitigating the incident upon arrival, while also enhancing their safety.
• A fire alarm notification might be based on a smoke-detector activation, which could have been triggered by an event as benign as food cooking on a stove; however, if a sensor in the home indicates a significantly higher temperature, particularly over a wide area, that would indicate a high probability of a structure fire.
• The type of response to a notification triggered by a carbon-dioxide detector would be impacted if the sensor was able to indicate the exact level of the gas that was present, recommending the need for emergency medical services to respond to treat victims with respiratory distress.
• PPVAR’s goal in developing the alarm-scoring system is to relay a great many discrete pieces of information as part of the alarm activation—which aren’t being relayed at all today—that will help determine whether the incident is legitimate and, if so, the type of response that needs to be sent. And the more relevant information that is provided to responders prior to incident arrival, the better positioned they will be to take appropriate action.

When such information is unavailable, the consequences can be tragic. We are aware of an instance when the son of a bank owner entered the building after closing to copy documents. Unfortunately, he did not enter the correct code into the alarm-system keypad and compounded his mistake by not closing the front door. When police officers arrived, they found the front door open, which was suspicious, and entered the building. They then confronted the young man and, mistaking him for a robber—which is what they were expecting because they had no information to the contrary—shot him to death. If the opportunity had existed to score the incident, and relevant information had been available to the responding officers—for instance, video that showed the young man making copies and not cleaning out the cash drawers—the tragedy likely would have been averted. The level of intelligence envisioned by PPVAR would keep both officers and the people they are sworn to protect safer.

Additional benefits exist as well beyond those provided to law-enforcement and fire/rescue/emergency medical agencies. For example, alarm-company clients will benefit from the expected significant drop in the number of false alarms. Most municipalities have ordinances that result in clients being charged a fee when their alarm systems generate more than three false alarms in a specified time, which could be monthly, quarterly, biannually, or annually, depending on the ordinance.

Given the scoring system envisioned by PPVAR, a client could arrange with its alarm company to receive a call for any notification up to a certain predefined scoring level; the client then could determine whether law enforcement needs to be called. If the score exceeds the predefined level, the notification would be sent immediately and automatically to the law-enforcement agency. Similarly, if the score is low enough, the alarm company might expand its effort (or take additional time) to get in touch with the client to confirm the type and severity of the incident, which in turn could prevent a call to law enforcement or fire/rescue/emergency medical. In any event, an approach such as this would enable the client to reduce false alarms dramatically, saving it significant money in the process.

Unverified alarms, particularly false alarms, represent a large problem for emergency-response agencies, including police departments and sheriff’s offices. The scoring system being developed by PPVAR will be a giant leap toward solving this problem.

Frank Fernandez is a PPVAR board member. He is president of Blueprints 4 Safety (B4S) Strategies Group located in Miami, Florida. An expert in police practices, he is a consultant to the United States Department of Justice, Civil Rights Division. He has 30-plus years of experience in law enforcement, beginning with the Miami Police Department rising to the rank of deputy chief, then serving as police chief in Hollywood, Florida. He retired after serving as public safety director and assistant city manager for Coral Gables, Florida.

John Chiaramonte also is a PPVAR board member. He is consulting division president of Mission Critical Partners, a public safety communications and information technology firm that supports PPVAR’s mission.