Disaster alerts reveal a better way to design tech
By Mark Hurst • September 15, 2023

This week on Techtonic I spoke with Jeannette Sutton, a professor at the University of Albany and an expert in disaster alerts. I was joined in the studio by WFMU’s station manager Ken Freedman, who has a lot of experience dealing with alerts, including the EAS (Emergency Alert System) that radio stations are required to test occasionally.

Disaster alerts are in the news recently, due to the wildfires in Maui and hurricane Hilary striking southern California. In Maui, wireless alerts – the smartphone popup message with the loud siren – didn’t reach enough residents in time. The New York Times described the problem in Maui Sent an Evacuation Alert. Why Did So Few People Get It? (by Mike Baker, Sergio Olmos and Eileen Sullivan, Sep 3, 2023):

“Evacuate your family and pets now, do not delay,” the warning said.

But many people most in need of the alert said the message never reached their phones, leaving them scrambling for safety as the fire began roaring toward their homes. More than 100 people died in the inferno, and some survivors wondered why they had not been notified earlier that the situation was out of control.

One of the challenges in Maui, as elsewhere, is that fewer people today tune in to the radio, opting instead for smartphones on a cell network:

As more households disconnect landline telephones and fewer families have access to broadcast television and radio, emergency managers have increasingly focused on wireless emergency notifications to instantly reach large numbers of people. But that modern system has its own limitations, relying on the resilience of cell networks and the proficiency of emergency crews across a patchwork of local agencies.

As Ken Freedman pointed out during the show, a cell tower only offers coverage within a relatively small area. If the disaster causes the power to go out, or if the tower is otherwise incapacitated, it becomes much harder to push an alert out to nearby smartphones. By comparison, broadcast radio is much more robust, as the towers can be 30 or 40 miles away and still offer a good signal to anyone with a radio.

And yet smartphones remain overwhelmingly popular. Even as radio and broadcast TV can and should broadcast alerts, any emergency plan has to include digital devices and platforms to maximize the reach of the message.

How to write an alert

On focus of Jeannette Sutton’s research is how emergency alerts should be written. Too often, she says, the emergency manager has to type the message into a blank text box, without any guidance or suggested phrases. The disaster – fire, hurricane, tornado, or whatever else – puts enough stress on the manager that it can be challenging to produce a message that is both accurate and complete.

On her site The Warn Room, Sutton lists the five parts of a good alert message: “Source, hazard name/description, location, consequences of threat, protective action, and time.” (She describes these in more detail during the interview.) Sutton is also working on a new software dashboard to guide emergency managers through the process of writing clear and accurate alert messages.

Lessons from disaster alerts

The conversation with Jeannette and Ken (listen to it here or see the episode page with links) touched on several lessons:

1. Older technology tends to have more robust infrastructure, and tends to be less widely used.

2. Newer tech tends to be more popular and more fragile.

3. Infrastructure is not the only consideration. The content of the messages also has to be carefully designed.

4. A good disaster alert plan will send the message across multiple platforms: broadcast radio, broadcast TV, wireless smartphone alerts, text messages, major social media platforms (notwithstanding their unethical practices), and possibly individual websites as well.

5. On the receiving end, citizens/users should have multiple devices to receive disaster messages: this must include a radio, as well as whatever digital devices (smartphone/laptop) they use.

Strategy matters

One thing I found refreshing about the disaster alerts conversation was the strategic thinking about technology. So often the talk about digital tech focuses on which platforms should be abandoned, and which are viable alternatives. (I’m pointing at myself here, as I constantly urge people to get off of the unethical Big Tech monopolies.) But in disaster planning, issues of infrastructure and message design are much more practical and nuanced, requiring consideration of multiple factors.

Sometimes I wonder how much strategic thinking is left in the tech industry – I mean strategy practiced for the good of a user base or a community. The Big Tech platforms are full of surveillance, dark patterns, and other monopoly tactics. Fewer teams seem to be invested in real customer experience strategy. This should change. As I wrote in Customers Included, the design of a digital product has to be treated as a strategic challenge, considering the needs of the users and the ability of the organization to deliver on those needs.

Are there still teams out there that engage in strategic thinking, or want to improve? Drop me a line if I can help.


An update on last week’s column, Don’t throw technology at it, in which I mentioned a news story about how human fire watchers, in remote regions, are being replaced by cameras and AI algorithms. In response, reader sent me the following, and with their permission I’m sharing it here:

Thought you’d like to know of a story I heard from an AI CEO at a startup which crashed & burned last summer. This was before the flameout, but has to do with flames, specifically, the Forest Service. The CEO told me that the Forest Service hired Palantir on the basis of its AI cred to use technology to monitor forests for fires. Turns out, that consists of sitting a human in front of a giant bank of TV monitors, like Watchmen’s Ozymandias, looking for smoke. Precious little tech involved: just cost-cutting. Replacing humans in watchtowers with one human watching video feeds from multiple watchtowers.

I don’t blame the Forest Service: especially under Trump, its budget was slashed. But I do wonder whether the fire detection is worse now and what money it saved. Probably some fractional salary – but not even maintenance on the towers & infrastructure since those cameras have to go somewhere. Sigh.

Once again we see the importance of thinking strategically about a problem, not just mindless throwing technology at it – or handing over one’s budget to a surveillance company.


An interview of me: Speaking of putting out fires, I was interviewed by the Modern Fire Instructor podcast about my 2007 book Bit Literacy and how it’s (still) relevant to teams with a mission to accomplish.

The episode is called “Turning Data Chaos into Productivity Tools,” listed on this page.

Thanks to host Rob Kandle, whose first question was about one of my favorite books, Robert Pirsig’s Zen and the Art of Motorcycle Maintenance. Interesting conversation.


You are missing out on most of my resources because you’re on a free subscription. On the Creative Good Forum, our members-only online community, you’re missing posts on . . .

• New entries in Good Experience Games (available only to Creative Good members), like new text-adventures and clever new in-browser games

• New entries in This Is Broken (also only for Creative Good members), such as a complaint about smart TVs

• New entries in Good Reports, such as alternatives to Google’s Chrome malware browser

• News about space junk, and Elon Musk’s misadventures, and Google’s antitrust trial, and a lot more.

Please support my work by joining Creative Good, and you’ll get access to the Creative Good Forum and all of its resources. Thanks.

Until next time,

-mark

Mark Hurst, founder, Creative Good – see official announcement and join as a member
Email: mark@creativegood.com
Listen to my podcast/radio show: techtonic.fm
Subscribe to my email newsletter
Sign up for my to-do list with privacy built in, Good Todo
On Mastodon: @markhurst@mastodon.social

- – -