The Planet Hunting Machine

Hundreds of inexpensively made antennae tuned into the cosmos look for radio waves from potentially habitable planets at Caltech’s Owens Valley Radio Observatory near Big Pine, California.
Hundreds of inexpensively made antennae tuned into the cosmos look for radio waves from potentially habitable planets at Caltech’s Owens Valley Radio Observatory near Big Pine, California.
Caltech professor Gregg Hallinan and his team are scanning the universe for possible Earths with cheap antennae built from parts found online, cake pans, and, yes, chicken wire. They’re taking the lonely search for habitable planets and automating it.

Sometime in the next decade, NASA hopes to deploy a rover to the dark side of the moon, where it will roll out 128 small, lightweight radio antennae in a flower configuration over 100 square kilometers of the lunar dirt. The FARSIDE project is designed to look for habitable planets in other solar systems.

But you can see it today, already in operation, here in California.

The Cosmic Dawn is happening again, so get your pressure suit on. To look for life “out there,” go visit Gregg Hallinan. He’s a professor of astronomy at the California Institute of Technology. And with about $70 worth of unleaded gas, you can drive from anywhere in the Golden State to his crisp-white shipping container in the Owens Valley and take a gander outside our solar system.

The Owens Valley is a spectacle of nature, and even more beautiful for being in the middle of nowhere. The Mammoth Mountain ski area looks down on it from the northwest. The 4,000-year-old bristlecone pines of the White Mountains do the same from the east. Los Angeles, to the south, still owns the valley floor, in order to suck the water its way (“Forget it, Jake. It’s Chinatown”). Locals from Bishop and Big Pine brag of the golden trout they catch in the rivers. It’s high desert, a xerophytic ecosystem fed by bajadas and alluvial fans that sprout next to nothing—bitterbrush, burrobush, buckwheat, creosote, and the occasional lizard or jackrabbit.

There’s a certain early-space-era design aesthetic that takes the breath away when you see it up close. At Caltech’s observatory in the Owens Valley, the 40 Meter radio telescope triggers that type of “drink your Tang” awe. The locals call it Big Ears. Nearly a million pounds of steel tubing and 14,000 square feet of aluminum paneling slathered in 1.5 tons of NASA-white paint can’t help but give you a 130-foot heartbeat. The 40 Meter has been in Hollywood films. It’s got movie star presence. Twice a week, it checks on about 1,800 blazars, which are jets of radioactive matter shooting at us from black holes. It has been the celebrity here for half a century. “The 40 Meter scope could pick up a cell phone on Pluto,” Mark Hodges, an OVRO design engineer, tells me.

But I’m not interested in the shining white celebrity. I’m here to find out about the cheap tricks being played across Caltech’s 1,200-acre property.

I’d heard that Hallinan had built low-cost ground antennae—some that used actual chicken wire, others that used cake pans ordered from Fat Daddio’s baking company in Spokane, Washington. Hooking them up to supercomputers, he had turned his hack into the most interesting telescope on the planet.

Caltech professor Gregg Hallinan examines the receiver from one of the antennae he designed for the Long Wavelength Array. His team can build an antenna in as little as 15 minutes.GORDON WILTSIE
Caltech professor Gregg Hallinan examines the receiver from one of the antennae he designed for the Long Wavelength Array. His team can build an antenna in as little as 15 minutes.

Regularly, the sun fires off coronal mass ejections, a billion tons of plasma hurled our way at more than a million miles per hour, Hallinan explains. Our atmosphere is not strong enough on its own to protect us, but it’s shielded by Earth’s magnetic field. The collision of coronal mass ejections and atmosphere sparks auroras at our poles.

I’ve followed Hallinan through a swing gate at the end of a dirt road into his antenna field with three others from his team. In the bright sun, the hundreds of square-tube aluminum antennae radiate white. They’re like tepee frames, about the same size as the bitterbrush that grows between them. When you look at them, their function isn’t readily apparent. They could just as easily be the work of an abstract artist as the work of an astronomer.

“To look for planets that could foster habitability in other solar systems, we can look for the pulse radio signal associated with this aurora emission,” says Marin Anderson, a postdoc on Hallinan’s team. “There are nearly 4,000 exoplanets”—planets orbiting a star other than our sun—“within about 75 light-years. We look for the signature of a magnetosphere.” She’s careful not to overpromise. “There are a lot of unknown unknowns. It might be below our sensitivity threshold.”

“We could have built this 40 years ago,” James Lamb, the observatory’s site director, chimes in. “We just didn’t have the computing power to stitch the signal together.” The 288 antennae here in the Owens Valley are not in a flower shape. They’re almost randomly placed.

“It looks random, but in fact the distance between each of the antennas is unique,” says Jonathon Kocz, a team member. “Radio waves from the universe hit each antenna at slightly different times.”

From this asynchronous pattern, the computers Kocz designed—called correlators—discern the precise direction the radio waves came from. The combination of distances between antennae gives Hallinan and his team 512 discrete measurements 200 million times per second. That’s 102.4 billion snippets of data every second to process and analyze in the quest for exoplanets.

The white shipping container that houses the supercomputers has to be kept between 17 and 20 degrees Celsius; any temperature variation beyond that makes its signals less stable, throwing its calibration out of whack. The air-conditioning unit is so big, its motor screams like a jet engine.

So far, Anderson has only scratched the surface—looking at 31 hours of data. Later this year, she and her teammates will upgrade the computer clusters and increase the number of antennae to 352; they hope to have thousands eventually. Within two years, the system will be running around the clock and they’ll be able to process 1,000 hours of data.

“The universe is our laboratory,” adds Anderson. “We just have to wait for the conditions that the universe sets up to play out.”

I imagined that looking for life on other planets was aspirational, lonely, impossibly poetic but implicitly futile. Worst of all, terracentric, which is to an astronomer what chauvinism might be to a women’s studies scholar. Damn, was I wrong.

Fundamentally, Hallinan and his team are working to take the solitary search for habitable planets and automate it. With less than a million dollars, they have made a planet-searching machine.

Known as the 40 Meter for the breadth of its diameter, this radio telescope is “so powerful that it could pick up a cell phone on Pluto,” says Mark Hodges, an OVRO design engineer.GORDON WILTSIE
Known as the 40 Meter for the breadth of its diameter, this radio telescope is “so powerful that it could pick up a cell phone on Pluto,” says Mark Hodges, an OVRO design engineer.

Hallinan is from County Sligo in Ireland. He’s a bundle of energy. Sort of a human pulsar. When he arrived at Caltech in 2012 as a young professor, the excitement in astronomy was at the high-energy end of the electromagnetic spectrum—gamma rays, UV rays. Huge, expensive telescopes that measured gravitational field events and black holes were the rage. But Hallinan was interested in ordinary radio waves, which are at the other end of the electromagnetic spectrum. Lower frequency than those from KPFK-FM, broadcasting from Los Angeles.

There’s almost no energy in radio waves. “All the energy from all the radio waves ever collected by all of the radio telescopes ever in existence wouldn’t be enough to lift a chicken feather a hundredth of an inch,” says Hodges, assuring me that he calculated an actual chicken feather’s weight to be sure his math was right.

While the telescopes of other universities were being launched into outer space or perched on spectacular mountaintops, Caltech’s radio telescopes were grounded. Literally and figuratively. In 2015, the National Science Foundation had put off funding and the Owens Valley Radio Observatory was down to six or seven staff members.

“It was a tough time,” Hallinan recalls. But he had a plan. “I wanted to work at factors of 10,000 times lower.”

Hallinan knew that big telescopes, for all their amazing capacity to bring a tiny speck of the sky into stunning focus, were helpless when it came to imaging the whole sky. He was going to go small. He invited some Caltech donors to Owens Valley and let one of them climb up the support leg of the movie star to the very tip of its receiver, in the center of the dish. Afterward, Hallinan explained his vision of capturing the entire sky.

He described to the patrons how, at every moment of every day, Earth is showered with infinitesimally weak radio waves generated by everything that has ever happened in the universe, whether it was 12 light-years ago or 12 billion. It’s called “the afterglow.” On their journey, radio waves are ever so slightly slowed down by free electrons, gases, plasma, and planets, and that provides very valuable information. Hallinan’s hunch was that signatures of other Earths could be found hidden in these zettabytes of daily radio data.

“People thought my idea was possible. They just didn’t believe I could do it for only a million dollars,” he says. The donors put in $100,000, which kick-started Caltech to follow on with another $450,000.

There’s a machine shop on the property stuffed with metal lathes, welders, and drill cutters. A legend of radio astronomy instrumentation design, Sandy Weinreb, cut roughly $3,500 off of Hallinan’s cost per antenna by custom-fabricating key parts and buying others online at a discount. Hallinan initiated a race between graduate students to assemble the antennae; one could be made in about 15 minutes. “It was great for morale,” he recounts. “But bad for execution.”

Lamb had to retighten their bolts afterward.

As the first 256 antennae were stabbed into the ground in 2013 and 2014, the team found cows using them to scratch themselves. Cheap fencing went up. Then chicken wire was laid on the ground, but it wasn’t to keep jackrabbits away; it bounces just a little radio wave back up to the underside of the antenna, refining the signal. “The chicken wire is pretty important,” Hallinan says.

In 2015, they added 32 more antennae—these ones spread out from the core group to make the array more sensitive—and by 2016, they had begun analyzing data. Hallinan’s Long Wavelength Array tuned into the galactic background, the radiation from cosmic rays zipping around the galaxy.

But this stuff, heavenly as it might be, wasn’t what Hallinan was really after. He puts it this way: “I like things that go bump in the night.”

When you tune in to the entire history of the universe, all at once, there’s a lot of math involved in separating out what’s always there from what’s event related—what radio astronomers lovingly call transients.

Deconvoluting it all to focus on transients of interest was the job of Anderson, Kocz, and others. Their computers weren’t looking for supernova remnants, the glow of the Milky Way, the North Galactic Spur. Their computers also ignored signal blasts from meaningless transient activity, such as meteorites hitting Earth’s atmosphere or a plane over Vegas reflecting a television station in New Mexico.

“Radio astronomy was extremely cool—back in the 1930s,” Vikram Ravi, a Caltech astronomy professor who’s researching fast radio bursts with Hallinan, says, laughing. “And in the ’60s and ’70s, it was used to find the first quasars and supernova remnants. But then astronomy went away from radio. Progress required bigger, multibillion-dollar telescopes.”

Just a month before I visited, Ravi helped discover a galaxy 7.9 billion light-years away, using a telescope with some parts machined in the observatory’s shop for $5,000—which included the cake pans. “But all of a sudden, because we’re building these telescopes by hand in the machine shop, they’re way more accessible,” he says. “We can do end-to-end engineering with as few as five people. It’s absolutely exciting a young generation.”

Hundreds of low-cost antennae from the Long Wavelength Array project work alongside conventional radio telescopes at the 1,200-acre Owens Valley property, beneath the eastern escarpment of California’s Sierra Nevada.GORDON WILTSIE
Hundreds of low-cost antennae from the Long Wavelength Array project work alongside conventional radio telescopes at the 1,200-acre Owens Valley property, beneath the eastern escarpment of California’s Sierra Nevada.

Here on Earth, searching for the colorful night lights of the aurora borealis is a bucket list adventure. Iceland, Norway, and Alaska get droves of aurora tourists a year. The odds of seeing it are small, and the chances this November are even smaller, because our sun is in the trough of an 11-year sunspot cycle.

Even though Hallinan and his team are looking for auroras light-years away, the probability of them spotting one is incredibly higher than if they were packing for Norway. They’ve removed the randomness of having to be in the right place at the right time.

If, like me, you thought Earth was special—strap in. It turns out that all the raw molecular ingredients of life are fully distributed throughout the universe, both inside and between solar systems, on every rocky mass, and in every interstellar gas and plasma. Water is nearly everywhere, as is carbon. When young stars churn out amino acids and project them into stellar winds—the very amino acids that make up DNA and RNA. Comets carry these amino acids between solar systems, and meteorites protect them from gamma rays. We’ve found RNA building blocks in solar systems 400 light-years away. Oceans on Saturn’s moons are brewing with nutrients. And the extreme conditions that scientists believe gave rise to unicellular organisms on Earth are an ordinary gaseous explosion in space. Bacteria can spore and survive for thousands, maybe even millions, of years. Even photosynthetic bacteria have survived space travel.

The first star-orbiting planet outside our solar system to be discovered was found by researchers in Geneva in 1995. Then NASA’s Kepler Space Telescope found others, and exoplanet discovery exploded; already, astronomers have located more than 4,000 such planets. The day before I was at Big Ears, an international team of researchers found 2 more: Teegarden b and Teegarden c, both just 12.5 light-years away. And NASA’s Transiting Exoplanet Survey Satellite, launched last year, will outdo Kepler. “The TESS Mission telescope is projected to find orders of magnitude more in just the next two years; it’s studying three million stars and is expected to find 14,000 exoplanets,” says Hallinan. According to NASA, more than 1,000 of these exoplanets—all within 75 light-years away—are potential targets for more precise observation. And 10 from this group will likely be very Earth-like: rocky, no more than twice our size, and in the habitable zone of a star. Across our whole galaxy, the number of exoplanets that likely share characteristics with Earth climbs into the billions.

Looked at this way, it becomes highly improbable that Earth is the only planet with biology on it.

In fact, the planet our country talks most about visiting—Mars—is a cautionary tale. Mars very likely had life on it several billion years ago. But it lost its magnetic field when its molten core died out. Gravity plummeted, and the solar wind blew away the planet’s atmosphere. Hammered by gamma rays, any life that existed on the surface of Mars would have disappeared from sight.

But Mars notwithstanding, with tens of thousands of new exoplanets on the verge of being photographed, astronomy has a new frontier. Someone needed to come up with a way to figure out which of these planets had a magnetic field and an atmosphere that shielded it from solar wind.

Which was Hallinan’s plan all along.

When we do find the first habitable planet, there will be no exact moment of discovery. Galaxial eurekas are an accumulation of layers of insight over months or years. A single transient blip in the data leads to a lot of conversation between astronomers about what it means. Other telescopes are brought in to look for confirmation. Skeptics have to be persuaded. Only later, in the media, do we invent the lie that a scientist “just discovered” Earth 2.0.

For instance, in July 2018, Hallinan and his graduate student Melodie Kao lit the media up with the “discovery” of radio waves from a possible rogue planet 20 light-years away that had a magnetic field thousands of times more powerful than Earth’s. But they’d first spotted it two years prior and had spent the next two years getting increasingly more accurate measurements. And in fact, astronomers had long known about this celestial sphere—they’d thought it was a failed star, a brown dwarf. What Hallinan and Kao detected were auroras at its poles, though what’s powering them remains a mystery. Earth’s auroras are caused by coronal mass ejections from the sun colliding with the upper layers of the atmosphere. While this rogue planet doesn’t have an orbiting star, it’s possible that, like Jupiter, its auroras are caused by an orbiting moon. In any case, a brown dwarf wouldn’t have an atmosphere. Auroras implied a planet.

So what are the odds that we find a planet that can sustain life? They’re now about the same odds that Christopher Columbus, sailing east, would find the Americas. It won’t even be a human who discovers the first one. It will be a computer program. With the help of some chicken wire.

But it takes a human to imagine the machine.

And that’s what I was struck by, more than anything else at Owens Valley. Hallinan’s project was 1 percent aluminum, 1 percent silicon, and 98 percent imagination. I was in awe of the staggering creativity it took to think that a blip in a faint radio transmission could become a machine to find other Earths.

Hallinan is no longer short on money. His early successes helped loosen the purse strings of the National Science Foundation and others. With $2.2 million from the NSF, his aurora scanner is that much closer to someday running around the clock, hunting for planets continuously.

The interest of NASA has been piqued. There’s a quite good chance that, sometime in the next decade, Hallinan’s array will be re-created on the dark side of the moon. “I just had dinner with NASA administrator Jim Bridenstine,” says University of Colorado Boulder astronomer Jack Burns, who was on the presidential transition team for NASA and constructed the space program for the incoming Trump administration. Burns has been envisioning a radio telescope on the moon since the 1980s. “But until Gregg built the LWA [Long Wavelength Array], we just didn’t have the killer app that would justify it.”

This spring, NASA challenged Hallinan, Burns, and the Jet Propulsion Laboratory, NASA’s famed R&D center, to figure out if they could build a 128-antenna array on the far side of the moon for a billion dollars or less. “Everyone involved got really excited when we realized the engineering was feasible,” says Burns. Jeff Bezos’s Blue Origin moon lander, Blue Moon, could take the package down to the lunar surface. Hallinan’s antennae would draw power from an isotope like plutonium. A satellite would relay the data back to Earth. “All of the technology and computer processing and correlations have been worked out by Gregg’s group,” Burns adds. “He presented in April, at a pre–Decadal Survey symposium, our 10-year plan. It was very well received by our astronomy peers.”

Since the moon has no ionosphere, the FARSIDE array could listen to even lower-frequency radio waves without interference. “The LWA is pushing the limits farther than they’ve ever been pushed before,” Anderson says. “But it’s still fundamentally limited.” This is important, because there’s a cruel, strange irony to the whole mission.

Despite all the things Hallinan’s earthbound array can listen to, including from billions of light-years ago—the one thing it can’t listen to is Earth itself. Our atmosphere blocks our own radio signal. And this goes for other Earths, too. The LWA will find exoplanets with strong magnetospheres—it will find solar systems where big Jupiter-like planets play a protective role for the rocky planets closer to the sun—but the more similar those rocky exoplanets are to Earth, the less likely it is that we’ll be able to detect them here on Earth.

To find a planet that’s truly like Earth, we have to leave Earth.

And what then?

“Well, we can’t get to them,” says Burns. “We can’t even get a probe there. The chemical-propulsion rockets we have now are fine to get around the solar system. But we would need leaps in propulsion technology.”

His words turn speculative, but he inevitably appeals to our restless nature. “It could happen by the end of the century. There’s a lot of ideas. People are working on them. There’s no limit to the imagination.”

Po Bronson’s writing honors include the American Association for the Advancement of Science Journalism Award. He wrote about Planet’s shoebox-size satellites for Alta, Issue 7.


Use of and/or registration on any portion of this site constitutes acceptance of our User Agreement (updated 1/1/20) and Privacy Policy and Cookie Statement (updated 1/1/20). Your California Privacy Rights. Do Not Sell My Personal Information. The material on this site may not be reproduced, distributed, transmitted, cached or otherwise used, except with the prior written permission of Alta.