Student and alumnus create earthquake warning system


Alumnus Ryan Logsdon (left) and masters student Sean Gibbons (right) created Lighthaus, a system that features earthquake sensors in individual homes to provide early warnings. (Emily Smith | Daily Trojan)

A  USC student and alumnus are heading to the Clinton Global Initiative University Conference this month after they teamed up to bring earthquake early warning detection kits to homes across Los Angeles.

Sean Gibbons, a student at the Marshall School of Business, met Viterbi School of Engineering alumnus Ryan Logsdon about a year ago on a platform meant to connect entrepreneurs and people with project ideas.

“[We] met on campus and he told me about this idea he had for a way to use artificial intelligence and deep learning to detect earthquakes,” Gibbons said. “I thought it was awesome, so I jumped on board and the rest is history.”

Logsdon, who graduated in 2015, is the chief technology officer for the project, titled Lighthaus. He had the idea to create the system after learning that earthquakes could be “outpaced” by technology. Despite not having a background in geophysics, Logsdon holds a master’s degree in computer science with a concentration in intelligent robotics. He wrote a simulation of the system in just three nights.

“I checked all my parameters and it didn’t make sense to me that you could do this and nobody is doing it,” Logsdon said.

Logsdon said that he then learned the United States Geological Survey is working on an early detection warning system, but that it is still in its testing phase.

The USGS system is a network of 400 high-quality ground motion sensors placed throughout California using the California Integrated Seismic Network, according to the USGS website.

Lighthaus will feature sensors in individual homes. The company intends to privatize the earthquake early warning system, according to Gibbons, who serves as the CEO for Lighthaus.

“It’s essentially a device that you’re going to put into your home and plug into your wall and that device acts as a sensor that in combination with other devices in the area, forms sort of a network of sensors,” Gibbons said. “All of those are streaming data back to our central server.”

The system is able to remain at a cost of around $30 by using off-the-shelf parts and employing what Logsdon called “error-prone” software.

“We take on less desirable hardware sensors and in exchange, we have to add in some sensors that work in real time so that way we can take the noise out of the filter,” Logsdon said. “You’d otherwise never be able to use these hardware for what we’re intending, so you pass it through some very simple filters and it pulls the data out that we require and that we need.”

Lighthaus, which recently incorporated, plans to give out 400 free sensors to homes around L.A. in order to test the system. The company plans to be subscription based, so those who have a paid subscription will receive warnings of up to 30 seconds.

Ron Silverton was a Viterbi Innovation Mentor to Logsdon and Gibbons and has continued to informally advise the duo. He was excited to work with Lighthaus because he believes it can positively impact many lives.

“They have a strong core team and a compelling product that I believe could make a real difference in helping to save lives,” Silverton wrote in an email to the Daily Trojan.

After winning several competitions that have been critical to launching Lighthaus, Logsdon and Gibbons are taking their project to the Clinton Global Initiative University Conference, which serves as a platform for undergraduate and graduate students to discuss innovative ideas. If the two raise the most money in their GoFundMe campaign, they will receive an award from former president Bill Clinton and his daughter Chelsea.  

Logsdon said he enjoys working on this problem because it’s fun.

“I don’t really care that it’s not the most attractive because it’s not my field,” Logsdon said. “My field is robotics, and if you can sum up robotics in one sentence or one word it’s ‘error’ or ‘how to deal with errors’ and that’s what this problem’s all about.”