At 3:07 pm last Thursday, inside Fire Station 6 in Menlo Park, Calif., an alarm sounded, followed by an announcement over the loudspeaker: “Earthquake warning. Severe shaking expected in 30 seconds. Bay doors opening….. Earthquake warning. Severe shaking expected in 20 seconds. Bay doors opening…”
An earthquake early warning system, recently installed in the newly remodeled fire station, had been triggered by a 7.8 magnitude earthquake detected some 64 kilometers away, in Bodega Bay, Calif. The system sprang into action, opening doors, turning on emergency lights, and shutting off gas burners as it counted down to the imminent arrival of the shockwaves.
But the firefighters and guests inside the station did not duck and cover; instead, they admired the technology in action, for this was a scheduled test of the latest installation of SkyAlert’s IoT-connected earthquake warning system.
“We can distribute an earthquake warning massively and instantaneously,” said SkyAlert founder and CEO Alejandro Cantu. “We provide an audible warning—so people know it is coming, how strong it will be, and how much time they have, so they can protect themselves.
“We can also protect assets—turn off gas valves, turn off machinery—things that would be hard to do in that situation by hand.”
In the case of a fire station, Cantu said, “We turn on emergency lights so the firefighters don’t get injured, turn off the gas so the firehouse doesn’t catch fire itself, and open the doors so the fire engines can get out.”
During the 1994 Northridge earthquake, he pointed out, “One of the doors of a fire station got jammed and it took four hours to open it. We are using those historical problems in designing our system.”
SkyAlert’s technology has been evolving for years in Mexico. Over time, the company extended what was initially a pager-based emergency alert system to a comprehensive earthquake alert system that includes a proprietary network of seismic sensors. Recently, Cantu’s efforts got the attention of researchers and investors in the U.S.—and an opportunity to give SkyAlert’s IoT technology far broader reach. It has been integrated with the U.S. Geographical Service’s (USGS’s) ShakeAlert, an earthquake warning system on the West Coast of the United States that started making its alerts available for public distribution in some areas late last year.
Born to worry about earthquakes
Cantu says he was born to worry about earthquake protection—literally. “Born three months after the 1985 Mexico City earthquake, I grow up in a city that had stories of earthquake tragedies everywhere,” he says. “I developed a driving purpose to prevent loss of life from earthquakes.”
His father was in the business of running pager networks (specifically, the Mexican arm of Skytel), so it was no surprise that Cantu majored in telecommunications engineering in college. He ended up adding in a minor in geological science. He started working in Skytel’s R&D group in 2009, developing an emergency mass communications system that would allow the company to send emergency alerts simultaneously to millions of Skytel users. Cantu says he immediately started trying to figure out how to bring earthquake early warnings into that system. He spun off SkyAlert in 2011 to focus on that part of the problem.
Initially, Cantu says, SkyAlert worked with data collected by earthquake sensors that had been installed and were being maintained by the Mexico City government.
“By 2013,” he says, “I realized that the network was…” Cantu hesitated, searching for the politic word. “Inconsistent,” he concluded.
Cantu went on an around-the-world hunt for better technology, finally purchasing earthquake sensors in Israel. By 2016, he had deployed a network of 120 sensors covering more than 850,000 square kilometers in Mexico. SkyAlert also brought on more than 400 corporate customers, including Walmart and WeWork in Mexico City, where the system triggers an announcement over the public address system urging people to take cover. The company also developed an earthquake warning app for smart phones; Cantu reports that some 3 million people in Mexico have downloaded it to date.
Meanwhile, he says, he kept an eager eye on the development of the ShakeAlert. Though he was not sure exactly how he would be able to tap into that system and the U.S. market, he was confident that doing so would be a good move for his company.
Another Mexico City earthquake puts the spotlight on warning systems
Then, in 2017, came another Mexico City earthquake.
“Our system gave a 20-second warning,” Cantu says. “We heard stories of people who received that warning and exited a building that collapsed behind them. It saved lives.”
SkyAlert’s EEW device Photo: SkyAlert
And it got a lot of international attention, he says, including a visit from a group of U.S. scientists. Richard Allen, a seismologist at the University of California, Berkeley, who has been working in the field of earthquake warning since 2001, was part of that delegation.
“Researchers in the U.S. had been working on an early warning system for quite a few years,” Allen says. “Mexico had the oldest public early warning system in operation. The earthquake gave us a chance to find out how it performed and, in particular, how the public used it. So a team of four of us-–two seismologists, and two social scientists—spent a week going around interviewing people about the warnings they got and how they responded. We got a lot of positive feedback about the SkyAlert system,” Allen says.
With that data in hand, the U.S. researchers set up a meeting with SkyAlert’s leaders.
“We hadn’t met them before,”Allen said. “There is a research community that works on early warning globally, but as a private company, they weren’t a part of that.”
The visit was eye opening in several ways. “[SkyAlert] is a private sector company, focused on delivering a product that people want and need,” Allen said. The U.S. researchers quickly figured out that:
“ShakeAlert is a much more sophisticated system, driven by a dedicated, seismic network infrastructure that is expensive and has been built over decades...but the [Mexican] private sector system was delivering alerts, [while] in the U.S., at that time, we were not. Private companies are focused on delivering a product; in the academic sector, everyone always wants to make it a little bit better.”
An accelerated move to California
A week later, Cantu says, he got a call from one of SkyAlert’s future investors, urging him to apply to Berkeley’s SkyDeck startup accelerator program.
He did, and the accelerator chose SkyAlert as part of a class of 20, due to start at the end of 2017.
“I had three weeks between the call telling me I was accepted and having to move,” Cantu said.
(Allen didn’t know anything about this new Berkeley connection until Cantu arrived and reached out to him. The seismologist has since signed on as an advisor to SkyAlert.)
It was a scramble, but absolutely worth it, Cantu says. Thanks to the accelerator, he was able to connect with the ShakeAlert team at USGS, and learn about the data available as well as the extent of the U.S. market for earthquake alert systems. By the end of 2018, SkyAlert had managed to upload the USGS data to its alert systems.
SkyAlert meets the Menlo Park Fire District
In early 2019, firefighters from Menlo Park, Calif., contacted ShakeAlert officials, seeking suggestions of a company to hire to integrate earthquake warnings into the district’s fire stations. SkyAlert was on the list.
“I got super excited,” Cantu says. “This was our first opportunity to test in the U.S., and it was such an amazing project. It employs everything we do, so makes a real case study.”
SkyAlert sold the Menlo Park Fire District on its technology—particularly, its ability to discriminate moderate shaking from severe earthquakes, the redundancy in its alert network, and the ease of integrating the technology into the fire stations’ control systems. The company started building the specialized IoT response systems for the project in January, held its first demo of the technology at Menlo Park Fire Station 2 in April, and followed up with last week’s demo at Menlo Park Station 6.
The fire station protection system cost about $20,000 to build, Cantu says. “Six years ago, this would have been a million-dollar project. It’s a great time to be working in early warning, with cloud computing providing sufficient computing resources; machine learning and artificial intelligence able to detect trends in this massive amount of data; powerful networks to move this data fast; and IoT creating a use case.”
SkyAlert isn’t the only company looking to use the ShakeAlert data. Early Warning Labs, Regroup, Varius, Everbridge, and others are also developing tools for companies and individuals based on the data. For startups, the focus is generally on systems for businesses looking to protect their workers and their assets. App-based alerts for individuals, while useful, aren’t going to bring in enough dollars to support a business, Allen says. That’s why Berkeley researchers are developing their own app, MyShake, so at least something is available to individual users.
In spite of all this activity, Cantu says, “it’s a race against time. The technology is doing its part. Now it’s about getting people to use the technology to save lives and dollars before the next big earthquake.”
Tekla S. Perry is a senior editor at IEEE Spectrum. Based in Palo Alto, Calif., she's been covering the people, companies, and technology that make Silicon Valley a special place for more than 30 years. An IEEE member, she holds a bachelor's degree in journalism from Michigan State University.