In the last decade, the Keck Observatory in Hawaii, one of the world’s most powerful telescopes, has spent hours staring at the night sky in search of exoplanets and accumulating huge amounts of data about potential new worlds elsewhere in the Milky Way.

But maybe, Nate Tellis wondered, Keck might have picked up something else along the way. Somewhere in all that data, could there be a signal from an intelligent civilization trying to reach Earth?

Tellis is a scientist at the University of California at Berkeley, where, as his LinkedIn biography puts it, he spends his days “trawling” astronomy datasets for statistical deviations, trying to figure out whether they’re actually extraterrestrial pings. He searches particularly for laser light, powerful pulses of photons that could be as short as a nanosecond. Tellis, along with astronomer Geoff Marcy, recently dug into the Keck archives for data from 5,600 stars, observed between 2004 and 2016. Tellis and Marcy built a laser-detecting computer algorithm to comb through all that recorded starlight—and the result, detailed in a recent study in the Astronomical Journal, is the largest survey of its kind in the field of optical-based searches for extraterrestrial life.

It didn’t find anything. So far, this has been par for the course when it comes to the search for extraterrestrial life, better known by the shorthand SETI. Astronomers first began using telescopes to look for potential alien communication in 1960, and they have been met with silence ever since.

“I think when you’re doing a SETI project, it’s very important not to get discouraged by a null detection,” Tellis said. “SETI has been in process for about 60 years, and it’s been non-detection after non-detection after non-detection.”

Astronomers and engineers have spent that time developing more powerful technology to conduct SETI surveys. The majority of SETI searches have relied on radio telescopes, which scour the skies for signals in the radio and microwave parts of the light spectrum. In the 1960s, “lasers were new, tricky, low-power devices; by contrast, radio technology had been developing for decades and was relatively mature,” according to a history from the SETI group at Harvard.

These days, lasers can outshine the sun, albeit in tiny pulses. But a tiny pulse—preferably more than one, to prove it’s not a fluke—is all it would take for a distant, advanced civilization to tell Earth “hey, we’re here!” If humans can get really good at sending radio and laser signals, the reasoning goes, maybe intelligent civilizations beyond Earth can, too—and then send them our way.

Unlike radio SETI, optical SETI looks for signals in the visible portion of the light spectrum. Lasers travel well over galactic distances. The light, concentrated into a narrow beam that can be 10 times as bright as the sun, would experience less interference from interstellar dust and gas than radio waves might. Laser emissions are also capable of carrying massive amounts of information. The network of cables at the bottom of the ocean is a collection of pulses of light, firing at high frequencies to transmit digital data and bring us the internet.

The dataset Tellis used for his study contained thousands of observations of stars as young as 200 million years and stars as old as nearly 10 billion years. Keck’s instruments collected millions of photons of light from these stars. What Tellis and his algorithm looked for were brief surges in photons. The first run of the data reported 5,000 potential candidates for mysterious laser beams, but they were eventually ruled out, explained away as emissions from stars’ outer layers, cosmic rays from our sun, or internal reflections from telescope instruments. Tellis got some firsthand Keck time to observe at least one target, KIC 8462852, a star about 1,500 light-years from Earth. In 2015, astronomers announced the Kepler space telescope had observed an unusual dimming of its light, which some believe could be caused by structures built by an advanced civilization around the star. The light emission observed from KIC 8462852 was the best candidate for an alien laser beam in the survey before it was ruled out.

The results may not have been surprising, but the method is noteworthy, says Jason Wright, an astronomer at Penn State University who contributed to some of the software code Tellis and Marcy used in the study. Recycling astronomical datasets that were produced for another purpose is pretty unusual, but it makes sense. There is strong competition among astronomers for observation time on the world’s best telescopes, and SETI proposals are usually low on the priority list.

“If you proposed to do a laser SETI study on Keck with thousands of hours, there’s nobody that will let you do it,” Wright said. Meanwhile, there’s plenty of astronomical datasets sitting around, waiting for a second look. One man’s trash is another man’s treasure, even in the search for life in the universe. Tellis, Wright said, “was digging through all the trash in case someone threw out a diamond.”

Tellis’s survey, like all SETI surveys, has its limitations. The data examined only some types of stars, in a specific wavelength range, and in Earth’s cosmic neighborhood. The telescope may not have been able to detect signals that were too faint or too bright, and too far away.

Optical SETI also depends on something beyond our control: a laser beam must first be aimed at Earth for it to be detected. Imagine another life-form on a distant world conducting the same kind of search, Tellis said. “If we had pointed our telescope at Earth at sort of the distance that we’ve been doing here, we wouldn’t have seen us,” he said, because Earth is not firing a laser beam into the universe as a beacon of its existence. Other worlds may not be, either.

“Every single one of those stars could have a New York City, a Paris, a London, and we would have no idea,” Tellis said.