How To Make Those Emergency Phone Alerts More Useful And Less Terrifying

New Yorkers had a particularly rude awakening on Monday morning, when phones across the city blared with the earsplitting tone of an emergency phone alert, which until now have been reserved for extreme weather notifications and Amber Alerts. But this time, phones displayed the following banner message—the first of its kind—urging people to look out for the suspect behind the bombings in NYC and New Jersey:

The alert—delivered at around 8 a.m. EST—was criticized, deservedly so, for a few reasons, including the vague directive (“see media for pic”) and the lack of a link or other identifying information beyond suspect Ahmad Khan Rahami’s name. Since this was an unprecedented use of the emergency alert in NYC, its efficacy is hard to parse: Though Rahami was tracked down a few hours later, the agency responsible for sending out the alert could not say for sure if it helped authorities make the arrest.

“I still have not heard anything specific about what role the alert may have played in ultimately catching the suspect, but we think that sharing the information across all possible city platforms was helpful as the suspect was identified within hours of the public being notified through a variety of channels,” NYC Emergency Management Department press secretary Nancy Silvestri told Fast Company. The NYPD, which put in the request for the alert, did not respond to requests for comment.

One thing, however, we can all agree on: Even if this type of alert delivers on getting the word out, the content needs work.

The mechanism behind the alerts in question is the Wireless Emergency Alerts (WEA) system, which is regulated by the Federal Emergency Management Agency (FEMA) and sends geo-targeted messages to mobile phones. That’s why (October 09, 2016)’s alert did not include a link or photo of Rahami: FEMA has imposed strict limitations on the technology, which means the alerts are limited to 90 text-only characters.

“It’s almost obvious to say, ‘Gee, shouldn’t the picture pop up with the alert message?’ That’s just what we’ve become accustomed to in our media-saturated, smartphone-carrying society,” says Bob Iannucci, a Carnegie Mellon professor doing research on improving the WEA system. “The fact is, WEA was engineered based on assumptions about what the network was 15 years ago.” When WEA was being developed, Iannucci says, it was “relatively straightforward to make WEA piggyback on the same sort of mechanisms that carry text messages.” In the time it took to get wireless carriers and public TV stations on board, the system became more dated (WEA has only been delivering messages since 2012).

This helps explain why FEMA also prohibits the inclusion of links in WEA messages. “The idea was not to overload the network with all of a sudden a million people going to find stuff,” notes Martin Griss, another CMU professor working on WEA research. “By not having a link, you’re not encouraging them to go poking around.” As Griss noted, this is a quaint notion when Google is now just one click away. People are going to search for more information regardless—so why not just provide it within the alert?

According to Motherboard, it’s not just FEMA that is concerned about overtaxing networks; even companies like Apple expressed concerns about “congesting wireless networks” in a filing to the Federal Communications Commission (FCC) earlier this year.

One update proposed by the FCC that is feasible in the short term is increasing the character count of the alerts form 90 characters to at least 280 characters, and possibly up to 360 characters. In studying the current state of the WEA system, the CMU team (which includes Iannucci, Griss, and their colleague Hakan Erdogmus) is advocating for a new interface that would better reflect how we consume information in this day—say, an app with a stamp of approval from Homeland Security ensuring the messages are authentic. “Once it is cast as an app, the possibilities of having it evolve more quickly and incorporate rich media and so on becomes something that’s much more realizable,” Iannucci said. WEA alerts could then be made to look more like what users are familiar with, like a web page.

The FCC will be re-evaluating current WEA rules—specifically, the character count and the exclusion of images and links—next week, on September 29, when it holds an open commission meeting.

Another improvement, Iannucci explained, would be to tweak the structure of alerts to easily update as a situation unfolds—a concept the researchers call situational awareness. During an earthquake, for example, there might be a slew of messages as the situation unfolds—maybe even some that conflict with previous messages. (In the case of (October 09, 2016)’s incident, the alert could have automatically updated once Rahami had been captured.)

“We studied a mechanism that would allow the phone to essentially digest this series of messages and only present the most current situational view,” Iannucci said. “The scenario is: A family is getting ready to leave their house. One of the parents pulls a phone out, looks at it and says, ‘Which way am I supposed to go?’ And the answer is there. They don’t have to sift through five, 10, 15 messages and, in a state of panic, try to figure out which is the most recent. Instead, the phone does that.” When the researchers tested a prototype with human subjects, they found that this approach was far simpler to understand than the existing pop-up message format.

The issue, it seems, is not technological limitations so much as rules imposed by the powers that be at FEMA and wireless carriers, who might protest changes to WEA’s current form. “The phone could do a lot, and the new app could do a lot,” Erdogmus said. “We proved that this is actually quite feasible from a tech perspective. But how you make even those small changes in the end-to-end, complex workflow is a matter of negotiation and policy.”