Optical Atomic Clocks May Enable Centimeter-Scale GPS

Walking into Jun Ye’slab at the University of Colorado Boulder is a bit like walking into an electronic jungle. There are wires strung across the ceiling that hang down to the floor. Right in the middle of the room are four hefty steel tables with metal panels above them extending all the way to the ceiling. Slide one of the panels to the side and you’ll see a dense mesh of vacuum chambers, mirrors, magnetic coils, and laser light bouncing around in precisely orchestrated patterns.

This is one of the world’s most precise and accurate clocks,and it’s so accurate that you’d have to wait 40 billion years—or three times the age of the universe—for it to be off by one second.

What’s interesting about Ye’s atomic clock, part of a joint venture between the University of Colorado Boulder and the National Institute of Standards and Technology (NIST), is that it is optical not microwave, like most atomic clocks. The ticking heart of the clock is the strontium atom, and it beats at a frequency of 429 terahertz, or 429 trillion ticks per second. It’s the same frequency as light in the lower part of the red region of the visible spectrum, and that relatively high frequency is a pillar of the clock’s incredible precision. Commonly available atomic clocks beat at frequencies in the gigahertz range, or about 10 billion ticks per second. Going from the microwave to the optical makes it possible for Ye’s clock to be tens of thousands of times as precise.

The startup Vector Atomic uses a vapor of iodine molecules trapped in a small glass cell as the ticking heart of its optical atomic clock. Will Lunden

One of Ye’s former graduate students, Martin Boyd, cofounded a company called Vector Atomic, which has taken the idea behind Ye’s optical-clock technology and used it to make a clock small enough to fit in a box the size of a large briefcase. The precision of Vector Atomic’s clock is far from that of Ye’s—it might lose a second in 32 million years, says Jamil Abo-Shaeer, CEO of Vector Atomic. But it, too, operates at an optical frequency, and it matches or beats commercial alternatives.

In the past year, three separate companies have developed their own versions of compact optical atomic clocks—besides Vector Atomic, there’s also Infleqtion, in Boulder, Colo., and QuantX Labs, based in Adelaide, Australia. Freed from the laboratory, these new clocks promise greater resilience and a backup to GPS for military applications, as well as for data centers, financial institutions, and power grids. And they may enable a future of more-precise GPS, with centimeter-positioning resolution, exact enough to keep self-driving cars in their lanes, allow drones to drop deliveries onto balconies, and more.

And even more than all that, this is a story of invention at the frontiers of electronics and optics. Getting the technology from an unwieldy, lab-size behemoth to a reliable, portable product took a major shift in mind-set: The tech staff of these companies, mostly Ph.D. atomic physicists, had to go from focusing on precision at all costs to obsessing over compactness, robustness, and minimizing power consumption. They took an idea that pushed the boundaries of science and turned it into an invention that stretched the possibilities of technology.

How does an atomic clock work?

Like any scientist, Ye is motivated by understanding the deepest mysteries of the universe. He hopes his lab’s ultraprecise clocks will one day help glean the secrets of quantum gravity, or help understand the nature of dark matter. He also revels in the engineering complexity of his device.

“I love this job because everything you’re teaching in physics turns out to matter when you’re trying to measure things at such a high-precision level,” he says. For example, if someone walks into the lab, the minuscule thermal radiation emanating from their body will polarize the atoms in the lab ever so slightly, changing their ticking frequency. To maintain the clock’s precision, you need to bring that effect under control.

Inside the briefcase-size optical atomic clock. A laser (1) shines into a glass cell containing atomic vapor (2). The atoms absorb light at only a very precise frequency. A detector (3) measures the amount of absorption and uses that to stabilize the laser at the correct frequency. A frequency comb (4) gears down from the optical oscillation in the terahertz to the microwave range. The clock outputs an ultraprecise megahertz signal (5). Chris Philpot

In an atomic clock, the atoms act like an extremely picky Goldilocks, identifying when a frequency of electromagnetic radiation they are exposed to is too hot, too cold, or just right. The clock starts with a source of electromagnetic radiation, be it a microwave oscillator (like the current commercial atomic clocks) or a laser (like Ye’s clock). No matter how precisely the sources are engineered, they will always have some variation, some bandwidth, and some jitter, making their frequency irregular and unreliable.

Unlike these radiation sources, all atoms of a certain isotope of a species—rubidium, cesium, strontium, or any other—are exactly identical to one another. And any atom has a host of discrete energy levels that it can occupy. Each pair of energy levels has its own energy gap, corresponding to a frequency. If an atom is illuminated by radiation of the exact frequency of one such gap, the atom will absorb the radiation, and the electrons will hop to the higher energy level. Shortly after, the atom will re-emit radiation as those electrons hop back down to the lower energy levels.

During clock operation, a maximally stable (but inevitably still somewhat broadband-jittery) source illuminates the atoms. The electrons get excited and hop energy levels only when the source’s frequency is just right. A detector observes how much of the radiation the atoms absorb (or how much they later re-emitted, depending on the architecture) and reports whether the incoming frequency is too high or too low. Then, active feedback stabilizes the source’s frequency to the atoms’ frequency of choice. This precise frequency feeds into a counter that can count the crests and troughs of the electromagnetic radiation—the ticks of the atomic clock. That stabilized count is an ultra-accurate frequency base—a clock, in other words.

There are a plethora of effects that can affect the precision of the clock. If the atoms are moving, the frequency of radiation from the atoms’ reference point is altered by the Doppler effect, causing different atoms to select for different frequencies according to their velocity. External electric or magnetic fields, or even heat radiating from a human, can tweak the atoms’ preferred frequency. A vibration can knock a source laser’s frequency so far off that the atoms will stop responding altogether, breaking the feedback loop.

Ye chose one of the pickiest atoms of them all, one that would offer very high precision—strontium. To minimize the noisemaking effects of heat, Ye’s team uses more lasers to cool the atoms down to just shy of absolute zero. To better detect the atoms’ signal, they corral the atoms in a periodic lattice—a trap shaped like an egg carton and made by yet another laser. This configuration creates several separate groups of atoms that can all be compared against one another to get a more precise measurement. All in all, Ye’s lab uses seven lasers of different colors for cooling, trapping, preparing the clock state, and detection, all defined by the atoms’ particular needs.

The lasers enable the clock’s astounding precision, but they are also expensive, and they take up a lot of space. Aside from the light source itself, each laser requires a bevy of optical control elements to coax it to the right frequency and alignment—and they are easily misaligned or knocked slightly away from their target color.

“The laser is a weak link,” Ye says. “When you design a microwave oscillator, you put a waveguide around it, and they work forever. Lasers are still very much more gentle or fragile.” The lasers can be knocked out of alignment by someone lightly knocking on one of Ye’s massive tables. Waveguides, meanwhile, being enclosed and bolted down, are much less sensitive.

The lab is run by a team of graduate students and postdocs, bent on ensuring that the laser’s instabilities do not deter them from making the world’s most precise measurements. They have the luxury of pursuing the ultimate precision without concern for worldly practicalities.

The mind-set shift to a commercial product

While Ye and his team pursue perfection in timing, Vector Atomic, the first company to put an optical atomic clock on the market, is after an equally elusive objective: commercial impact.

“Our competition is not Jun Ye,” says Vector Atomic’s Abo-Shaeer. “Our competition is the clocks that are out there—it’s the commercial clocks. We’re trying to bring these more modern timekeeping techniques to bear.”

To be commercially viable, these clocks cannot be thrown off by the bodily heat of a nearby human, nor can they malfunction when someone knocks against the device. So Vector Atomic had to rethink the whole construction of its device from the ground up, and the most fragile part of the system became the company’s focus. “Instead of designing the system around the atom, we designed the system around the lasers,” Abo-Shaeer says.

First, they drastically reduced the number of lasers used in the design. That means no laser cooling—the clock has to work with atoms or molecules in their gaseous state, confined in a glass cell. And there is no periodic lattice to group the atoms into separate clumps and get multiple readings. Both of these choices come with hits to precision, but they were necessary to make robust, compact devices.

Then, to choose their lasers, Abo-Shaeer and his coworkers asked themselves which ones were the most robust, cheap, and well-engineered. The answer was clear—infrared lasers used in mature telecommunications and machining industries. Then they asked themselves which atom, or molecule, had a transition that could be stimulated by such a laser. The answer here was an iodine molecule, whose electrons have a transition at 532 nanometers—conveniently, exactly half the wavelength of a common industrial laser. Halving the wavelength could be achieved by means of a common optical device.

“We have all these Ph.D. atomic physicists, and it takes as much or more creativity to get all this into a box as it did when we were graduate students with the ultimate goal of writing Nature and Science papers,” Abo-Shaeer says.

Vector Atomic couldn’t get away with just one laser in its system. Having a box that outputs a very precise laser, oscillating at hundreds of terahertz, sounds cool but is completely useless. No electronics are capable of counting those ticks. To convert the optical signal into a friendly microwave one, while keeping the original signal’s precision, the team needed to incorporate a frequency comb.

Frequency combs are lasers that emit light in regularly spaced bursts in time. Their comblike nature becomes apparent if you look at the frequencies—or colors—of the light they emit, regularly spaced like the teeth of a comb. The subject of the 2005 Nobel Prize in Physics, these devices bridge the optical and microwave domains, allowing laser light to “gear down” to lower frequency range while preserving precision.

In the past decade, frequency combs underwent their own transformation, from lab-based devices to briefcase-size commercially available products (and even quarter-size prototypes). This development, as much as anything else, unleashed a wave of innovation that enabled the three optical atomic clocks and this nascent market today.

High time for optical time

Inventions often happen in a flurry, as if there were something in the air making conditions ripe for the new innovation. Alongside Vector Atomic’s Evergreen-30 clock, Infleqtion and QuantX Labs have both developed clocks of their own in short order. Infleqtion has made seven sales to date of their clock, Tiqker (yes, quantum-tech companies are morally obligated to put a q in every name). QuantX Labs, meanwhile, has sold the first two of their Tempo clocks, with delivery to customers scheduled before the end of this year, says Andre Luiten, cofounder and managing director of QuantX Labs. (A fourth company, Vescent, based in Golden, Colo., is also selling an optical atomic clock, although it is not integrated into a single box.)

Vector Atomic, QuantX Labs, and Infleqtion all have plans to send prototypes of their clocks into space. QuantX Labs has designed a 20-liter engineering model of their space clock [left]. QuantX Labs

Independently, all three companies have made surprisingly similar design choices. They all realized that lasers were the limiting factor, and so chose to use a glass cell filled with atomic vapor rather than a vacuum chamber and laser cooling and trapping. They all opted to double the frequency of a telecom laser. Unlike Vector Atomic, Infleqtion and QuantX Labs chose the rubidium atom. The energy gap in rubidium, around 780 nm, can be addressed by a frequency-doubled infrared laser at 1,560 nm. QuantX Labs stands out for using two such lasers, very close to each other in frequency, to probe through a clever two-tone scheme that requires less power. They all managed to fit their clock systems into a 30-liter box, roughly the size of a briefcase.

All three companies took great pains to ensure that their clocks will operate robustly in realistic environments. At the lower level of precision compared with lab-based optical clocks, the radiation coming from a nearby person is no longer an issue. However, by doing away with laser cooling, these companies have heightened the possibility that temperature and motion could affect the atoms’ internal ticking frequency.

“You’ve got to be smart about the way you make the atomic cell so that it’s not coupled to the environment,” says Luiten.

Optical clocks set sail and take flight

In mid-2022, to test the robustness of their design, Vector Atomic and QuantX Labs’ partners in its venture, the University of Adelaide and Australia’s Defence Science and Technology Group, took their clocks out to sea. They brought their clocks to Pearl Harbor, in Hawaii, to participate in the Alternative Position, Navigation and Timing Challenge at Rim of the Pacific, a defense collaboration among the Five Eyes nations—Australia, Canada, New Zealand, the United Kingdom, and the United States. “They were playing touch rugby with the New Zealand sailors. So that was an awesome experience for atomic physicists,” Abo-Shaeer says.

After 20 days aboard a naval ship, Vector Atomic’s optical clocks maintained a performance that was very close to that of their measurements under lab conditions. “When it happened, I thought everyone should be standing up and shouting from the rooftops,” says Jonathan Hoffman, a program manager at the U.S. Defense Advanced Research Projects Agency (DARPA), which cofunded Vector Atomic’s work. “People have been working on these optical clocks for decades. And this was the first time an optical clock ran on its own without human interference, out in the real world.”

Vector Atomic and QuantX affiliate University of Adelaide installed their optical atomic clocks on a ship [top] to test their robustness in a harsh environment. The performance of Vector Atomic’s clocks [bottom] remained basically unchanged despite the ship’s rocking, temperature swings, and water sprays. The University of Adelaide’s clock degraded somewhat, but the team used the trial to improve their design. Will Lunden

The University of Adelaide’s clock did suffer some degradation at sea, but a critical outcome of the trial was an understanding of why that happened. This has allowed the team to amend its design to avoid the leading causes of noise, says Luiten.

In May 2024, Infleqtion sent its Tiqker clock into flight, along with its atom-based navigation system. A short-haul flight from MoD Boscombe Down, a military aircraft testing site in the United Kingdom, carried the quantum tech along with the U.K.’s science minister, Andrew Griffith. The company is still analyzing data from the flight, but at a minimum the clock has outperformed all onboard references, according to Judith Olson, head of the atomic clock project at Infleqtion.

All three companies are working on yet more compact versions of their clocks. All are confident they will be able to get their briefcase-size boxes down from a volume of about 30 liters to 5 L, about the size of an old-school two-slice toaster, say Olson, Luiten, and Abo-Shaeer. “Mostly those boxes are still empty space,” Luiten says.

During the sea trials, Vector Atomic’s and the University of Adelaide’s clocks were exposed to the elements. Jon Roslund

Infleqtion also has designs for an even smaller, 100-mL version, which leverages integrated photonics to make such tight packaging possible. “At that point, you basically have a clock that can fit in your pocket,” says Olson. “It might make a very warm pocket after a while, because the power draw will still be high. But even with the large power draw, that’s something we perceive as being potentially extremely disruptive.”

All three companies also plan to send their designs into space, aboard satellites, in the next several years. Under their Kairos mission, QuantX will launch a component of their Tempo clock into space in 2025, with a full launch scheduled for 2026.

Precision timing today

So why would someone need the astounding precision of an optical atomic clock? The most likely immediate use cases will be in situations where GPS is unavailable.

When most people think of GPS, they picture that blue dot on a map on their smartphone. But behind that dot is a sophisticated network of remarkable timing devices. It starts with Coordinated Universal Time (UTC), the standard established by averaging together about 400 atomic clocks of various kinds all over the world.

“UTC is known to be some factor of 1 million more stable than any astronomical sense of time provided by Earth’s rotation,” says Jeffrey Sherman, a supervisory physicist at NIST who works on maintaining and improving UTC clocks.

UTC is transmitted to satellites in the GPS network twice a day. Each satellite carries an onboard clock of its own, a microwave atomic clock usually based on rubidium. These clocks are precise to about a nanosecond during that half-day they are left to their own devices, Sherman says. From there, satellites provide the time for all kinds of facilities here on Earth, including data centers, financial institutions, power grids, and cell towers.

Precise timing is what allows the satellites to locate that blue dot on a phone map, too. A phone must connect to three or more GPS satellites and receive precise time from all three. However, the times will be different due to the different distances traveled from the satellites. Using this difference, and knowing the positions of the satellites, the phone triangulates its own position. So the precision of timing aboard the satellites directly relates to how precisely the location of any phone can be determined—currently about 2 meters in the nonmilitary version of the service.

The precisely timed future

Optical atomic clocks can usefully inject themselves into multiple stages of this worldwide timing scheme. First, if they prove reliable enough over the long term, they can be used in defining the UTC standard alongside—and eventually instead of—other clocks. Currently, the majority of the clocks that make up the standard are hydrogen masers. Hydrogen masers have a precision similar to that of the new portable optical clocks, but they are far from portable: They are roughly the size of a kitchen refrigerator and require a room-size thermally and vibrationally controlled environment.

“I think everyone can agree the maser is probably at the end of its technological evolution,” Shermann says. “They’ve stopped really getting a lot better, while on day one, the first crop of optical clocks are comparable. There’s a hope that by encouraging development, they can take over, and they can become much better in the near future.”

The global timing infrastructure. A collection of precise clocks, including hydrogen masers and atomic clocks, is used to create Coordinated Universal Time (UTC). A network of satellites carries atomic clocks of their own, synced to UTC on a regular basis. The satellites then send precise timing to data centers, financial institutions, the power grid, cell towers, and more. Four or more satellites are used to determine your phone’s GPS position. An optical atomic clock can be included in UTC, sent aboard satellites, or used as backup in data centers, financial institutions, or cell towers. Chris Philpot

Second, optical clocks can come in handy in situations where GPS isn’t available. Although many people experience GPS as extremely reliable, jammed or spoofed GPS is very common in times of war or conflict. (To see a daily map of where GPS is unavailable due to interference, check out gpsjam.org.)

This is a big issue for the U.S. Department of Defense. Not having access to GPS-based time compromises military communications. “For the DOD, it’s very important that we can put this on many, many different platforms,” DARPA’s Hoffman says. “We want to put it on ships, we want to put it on aircraft, we want to put it on satellites and vehicles.”

It can also be an issue in financial markets, data centers, and 5G communications. All of these use cases require precise timing to about 1 microsecond to function properly and meet regulatory requirements. That means the source of timing for these applications must be at least an order of magnitude better, or roughly a 100-nanosecond resolution. GPS provided this with room to spare, but if these industries rely solely on GPS, jamming or spoofing puts them at great risk.

A local microwave atomic clock can provide a backup, but these clocks lose several nanoseconds a day even in controlled-temperature environments. Optical atomic clocks can provide these industries with security, so that even if they lose access to GPS for extended periods of time, their operations will continue unimpeded.

“By having this headroom in performance, it means that we can trust how well our clocks are ticking hours and days or even months later,” says Infleqtion’s Olson. “The lower-performing clocks don’t have that.”

Finally, portable optical atomic clocks open up the possibility of a future where the entire timing fabric goes from nanosecond to picosecond resolution. That means sending these clocks into space to form their own version of a more-precise GPS. Among other things, this would enable location precision that’s several millimeters instead of 2 meters.

“We call it GPS 2.0,” says Vector Atomic’s Abo-Shaeer. He argues that millimeter-scale location resolution would allow autonomous vehicles to stay in their lanes, or make it possible for delivery drones to land on a New York City balcony.

Perhaps most exciting of all, this invention promises to open the possibility for many inventions in a variety of fields. Having the option of superior timing will open new applications that have not yet been envisioned. “A lot of applications are built around the current limitations of GPS. In other words, it’s sort of a catch-22,” says David Howe, leader of the time and frequency metrology group of NIST. “You get into this mode where you don’t ever cross over to something better because the applications are designed for what’s available. So, it’ll take a larger vision to say, ‘Let’s see what we can do with optical clocks.’”