News
Emergency notification extends its reach

During 9/11, for example, when the World Trade Center was collapsing,
no one in New York could make a cell phone call because the cell towers
were overloaded. The design of an emergency communications system is
critical — otherwise it could fail when it’s most needed.

Shortly after the shootings at Virginia Tech, in which the shooter
killed 32 people before shooting himself, the University of Toronto
decided to roll out its own emergency notification system. The
university chose a hosted solution from a Canadian company called Aizan
Technologies, since they were concerned about storing students’ and
faculty’s private data south of the border where it could be subject to
the U.S. Patriot Act. It allows the crisis management team to send out
a short message — a voice message, text-to-voice message or SMS — to
any phone number registered in the system.

“In a public school where you can lock the building down, you know
basically everybody’s in a classroom at 10:30am, whereas here there are
so many buildings,” says Erin Lemon, director of faculty and staff
communications with the U of T. “But everybody carries a cell phone.”

At this point, all landline phone numbers at the U of T have been
loaded into the system, but students have to subscribe to receive
text-to-voice alerts on their cell phone. “The research we did showed
us that you get much more reliable data that way than if you collect
everybody’s cell phone number at registration,” she says. When students
sign up for the service, they can also specify which of the three
campuses they spend the most time at, so messages can be targeted
appropriately.

The subscription service became available a few months ago. So far,
about 7,000 students have signed up (there are 70,000 students at the U
of T, and another 12,000 faculty and staff members). “We’re sitting at
about 10 per cent of our student population at this point,” says Lemon.
“The next step is to do a big push when the students come back in the
fall.” But the university doesn’t anticipate all students will sign up
for the service. If, say, three students out of 15 get the message in a
study group at the library, everybody will get the message.

The university can also send out a more detailed e-mail alert to
everybody with a U of T e-mail address in about an hour, which is much
faster than in the past. The system can be configured to stack the
calls, so they’re not overwhelming the cell towers. The university is
also experimenting with Facebook and Twitter, but at this point they
aren’t using social networking as part of the mix for emergency
messaging — though they could revisit that in the future.

“We wanted to keep the system quite clean so it’s only for
emergencies,” says Lemon. “Some schools will also alert you when your
exam schedule is up, but we want it to be like the fire alarm — when
you hear the fire alarm, you know exactly what to do.”

Whether it’s an on-premise or a hosted solution, organizations should
be looking at multi-modal alerting — by phone, by e-mail, by SMS — and
not relying on one single path, says Jayanth Angl, senior research
analyst with Info-Tech Research Group. There may be a situation where
phone lines are down or e-mail systems are offline or overburdened; or,
perhaps cellular infrastructure is at maximum capacity and unable to
transmit a text message in a timely fashion.

“SMS is an obvious mechanism here,” he says, “but the SMS service was
not really designed for emergency notification, and a surge in traffic
can affect the stability and operation of that network infrastructure.”
No single modality is 100 per cent guaranteed. Using a combination of
different modalities will ensure as many individuals as possible are
alerted as quickly as possible.

Twitter as a tool for communications and collaboration has a few pros
in that it’s multi-platform, so messages can be delivered via the Web
or via a mobile device. The concern, however, is it’s a passive model.
It has to be pushed to users — it’s a best-effort service, with no
service-level agreements attached to it. It’s been prone to failures
and outages, and there’s no ability to verify who’s been alerted. It
can be invaluable as an informational tool, says Angl, but in terms of
public safety and emergency notification, it lacks the robustness that
organizations would be looking for.

Organizations may want to consider integration with on-premise
elements, such as building security systems, paging systems and two-way
radios. If there’s a large user base distributed across multiple
locations, however, a hosted provider — which uses its global
infrastructure to deliver emergency notification — may be better suited
to that alerting function.

There are no codes or standards that are legally binding in Canada —
perhaps because the concept of having to notify so many people at once
is still relatively new. But, if you look south of the border, a number
of codes can provide some direction. The U.S. military, for example,
uses United Facilities Criteria (UFC). Every U.S. government facility
around the globe will have an emergency notification system in place by
the end of 2010. And the National Fire Protection Association (NFPA)
has been revising NFPA-72 to include a new section on emergency
communications.

According to the National Building Code of Canada and certain
Underwriters’ Laboratories of Canada (ULC) standards, a fire alarm
system with a voice component must be audible in every nook and cranny
of every building. A PA system, on the other hand, by law doesn’t have
to. So, while some codes don’t relate directly to emergency
communications, by their nature they’re able to provide best practices
in this area.

But it’s not enough to provide users with an alert — the system should
be capable of addressing key questions and concerns, and ensure users
have all the necessary information to take appropriate actions. This
may include different messaging for a person in facilities management
versus a general employee or constituent.

While two-way communications eat up network bandwidth, it’s critical to
know who didn’t get an alert, says Ken Dixon, executive vice-president
of worldwide sales with MIR3, which offers on-premise, hosted and
hybrid solutions for emergency communications, with six data centres,
including one in Canada. If you send out 2,000 alerts, and 200 people
don’t respond, you can quickly regroup those 200 people and start
sending follow-up notifications.

That’s the problem with Twitter — it doesn’t offer two-way
communications (though the company will be integrating Twitter and
other social networking modalities into its solution by year’s end).
Organizations have to consider whether it’s more important to get the
word out or know who didn’t get the word. “In the active shooter
environment, my concern is not getting the word to students off campus,
so they drive into the situation,” says Dixon.

The benefit of a hosted solution is that you’re not tied to your own
infrastructure. The next big incident after Virginia Tech was when the
Queens campus of St. John’s University in New York was locked down
after police apprehended a masked man with a rifle on campus. “We
called 17,000 students in less than two minutes,” says Dixon. “We buy
infrastructure and dedicate it to people — it’s not a physical
assignment, it’s a virtual assignment, so we don’t put more than two
people on a port.”

Most organizations are moving away from the older auto-dialer systems,
says Nicholle McClelland, marketing director with Blackboard Connect,
which offers hosted emergency notification for education and government
customers. “They’re looking for a system that’s hosted via the Web.”
Auto-dialers literally dial through all the phone numbers in the
system, so they can’t initiate messages in a timely fashion. And
keeping up with hardware requirements can be expensive and
time-consuming for staff. The company has service-level agreements with
telcos, where they’ve dedicated a certain amount of capacity to
initiate messages, so that’s guaranteed — however, there can be telco
errors or network congestion on a local level.

If you’re asking people to evacuate, you should also give them
instructions on where to go and what to do. You want to let them know,
for example, that it’s not safe to return to their campus dorm rooms.
“Just sending one message can create chaos because people don’t know
what to do,” says McClelland. Twitter is too passive for this. “For
urgent situations, you don’t want to wait for that tweet to come
through. It’s really best to hear [about it] from the principal.”

Universities are also turning to digital signage, since it gives them a
secondary medium to alert people in common areas. Several fire alarm
manufacturers have adopted LED technology for digital signage, so the
fire alarm panel can push an emergency message to displays.
Universities can also use the displays for non-emergency messages, such
as “football practice at 2pm today,” using supplementary connections
through the IT network. This trains students to read the displays on a
regular basis. UFC specifications call for amber alert strobes —
similar to strobe lights — to be placed adjacent to LED displays, so
people in a noisy area would see flashing yellow lights to alert them
to the message.

The faster your response, the faster your recovery. “Everything is
about response,” says Rod Piukkala, marketing director for the public
sector with Telus Communications, which offers the SafetyNet umbrella
of products including wireless emergency notification and crisis
management conferencing. This will help minimize the impact of an
incident, whether it’s financial or brand-related.

Organizations should consider their redundancies. Chances are there’s
going to be restricted travel after an incident, he says, but you still
have to be able to continue business. The better prepared you are to
continue business, the less recovery you’ll need. Organizations could
use audio, video and Web conferencing, for example, to share
information with other stakeholders, such as parents or suppliers.

“The process is the key element, because it makes people think outside
the box no matter what happens,” says Piukkala. “If you look at 9/11,
nobody ever thought a plane was going to fly into the World Trade
Center. Nobody had any distinct plans for that. So you have to think in
general terms.”