How an Atomic Clock Will Get Humans to Mars on Time


June 14, 2019


The Deep Space Atomic Clock, a new technology from NASA’s JPL, may change the way spacecraft navigate in space. Launching in late June aboard the Orbital Test Bed satellite, on the SpaceX’s Falcon Heavy rocket, descendants of the technology demonstration could be a key component of a self-driving spacecraft and a GPS-like navigation system at other worlds.

Credits: General Atomics Electromagnetic Systems

NASA navigators are helping build a future where spacecraft could safely and autonomously fly themselves to destinations like the Moon and Mars.

Navigators today tell a spacecraft where to go by calculating its position from Earth and sending the location data to space in a two-way relay system that can take anywhere from minutes to hours to deliver directions. This method of navigation means that no matter how far a mission travels through the solar system, our spacecraft are still tethered to the ground, waiting for commands from our planet.

That limitation poses obvious problems for a future crewed mission to another planet. How can astronauts navigate far from Earth if they don’t have immediate control over where they’re going? And how can they accurately land on another planet when there’s a communication delay that affects how quickly they can adjust their trajectory into the atmosphere?

NASA’s Deep Space Atomic Clock is a toaster-size device that aims to answer those questions. It’s the first GPS-like instrument small and stable enough to fly on a spacecraft. The technology demonstration enables the spacecraft to know where it is without needing to rely on that data from Earth. In late June, the clock will launch on the SpaceX Falcon Heavy rocket into Earth’s orbit for one year, where it will test whether it can help spacecraft locate themselves in space.

If the Deep Space Atomic Clock’s trial year in space goes well, it could pave the way for a future of one-way navigation in which astronauts are guided by a GPS-like system across the surface of the Moon or can safely fly their own missions to Mars and beyond.

“Every spacecraft exploring deep space is steered by navigators here on Earth. Deep Space Atomic Clock will change that by enabling onboard autonomous navigation, or self-driving spacecraft,” said Jill Seubert, the deputy principal investigator.

There’s No GPS in Deep Space

Atomic clocks in space aren’t new. Every GPS device and smartphone determines its location via atomic clocks on satellites orbiting Earth. The satellites send signals from space, and the receiver triangulates your position by measuring how long the signals take to reach your GPS.

Currently, spacecraft flying beyond Earth’s orbit don’t have a GPS to find their way through space. Atomic clocks on GPS satellites aren’t accurate enough to send directions to spacecraft, when being off by even less than a second could mean missing a planet by miles.

Instead, navigators use giant antennas on Earth to send a signal to the spacecraft, which bounces it back to Earth. Extremely precise clocks on the ground measure how long it takes the signal to make this two-way journey. The amount of time tells them how far away the spacecraft is and how fast it’s going. Only then can navigators send directions to the spacecraft, telling it where to go.

“It’s the same exact concept as an echo,” said Seubert. “If I’m standing in front of a mountain and I shout, the longer it takes for the echo to come back to me, the farther away the mountain is.”

Two-way navigation means that no matter how deep into space a mission goes, it still has to wait for a signal carrying commands to cross the vast distances between planets. It’s a process made famous by Mars landings like Curiosity, when the world waited 14 long minutes with mission control for the rover to send the message that it landed safely. That delay is an average wait time: Depending on where Earth and Mars are in their orbits, it can take anywhere from 4 to 20 minutes for a one-way signal to travel between planets.

It’s a slow, laborious way to navigate in deep space, one that ties up the giant antennas of NASA’s Deep Space Network like a busy phone line. During this exchange, a spacecraft flying at tens of thousands of miles per hour could be in an entirely different place by the time it “knows” where it is.

A Better Way to Navigate

An atomic clock small enough to fly on a mission but precise enough to give accurate directions could eliminate the need for this two-way system. Future navigators would send a signal from Earth to a spacecraft. Like its Earthly cousins, the Deep Space Atomic Clock onboard would measure the amount of time it took that signal to reach it. The spacecraft could then calculate its own position and trajectory, essentially giving itself directions.

“Having a clock onboard would enable onboard radio navigation and, when combined with optical navigation, make for a more accurate and safe way for astronauts to be able to navigate themselves,” said Deep Space Atomic Clock Principal Investigator Todd Ely.

This one-way navigation has applications for Mars and beyond. DSN antennas would be able to communicate with multiple missions at a time by broadcasting one signal into space. The new technology could improve the accuracy of GPS on Earth. And multiple spacecraft with Deep Space Atomic Clocks could orbit Mars, creating a GPS-like network that would give directions to robots and humans on the surface.

“The Deep Space Atomic Clock will have the ability to aid in navigation, not just locally but in other planets as well. One way to think of it is as if we had GPS at other planets,” said Eric Burt, the ion clock development lead.

Burt and fellow JPL clock physicists Robert Tjoelker and John Prestage created a mercury ion clock, which maintains its stability in space in the same way as refrigerator-size atomic clocks on Earth. In lab tests, the Deep Space Atomic Clock proved to be 50 times more accurate than GPS clocks. That’s an error of 1 second every 10 million years.

The clock’s demonstration in space will determine whether it can remain stable in orbit. If it does, a Deep Space Atomic Clock could fly on a mission as early as the 2030s. The first step toward self-driving spacecraft that could one day carry humans to other worlds.

The Deep Space Atomic Clock is hosted on a spacecraft provided by General Atomics Electromagnetic Systems of Englewood, Colorado. It is sponsored by the Technology Demonstration Missions program within NASA’s Space Technology Mission Directorate and the Space Communications and Navigations program within NASA’s Human Exploration and Operations Mission Directorate. JPL manages the project.

Here’s five things to know about NASA’s Deep Space Atomic Clock:

Learn about the other NASA missions on the SpaceX Falcon Heavy launch that’s carrying the Deep Space Atomic Clock:

0 0 votes
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
June 16, 2019 2:41 am

As long as they don’t make the mistake of interpreting measured meters as feet which wrecked a orevious mars mission. They switched off the braking rockets too soon and crashed on the martian surface.

Dan Hawkins
Reply to  Hans Erren
June 16, 2019 9:32 am

So the phrase “…a orevious mars mission” would make sense if I interpret it in metric?

Reply to  Dan Hawkins
June 16, 2019 1:30 pm

@#€&€ ipad keyboard

Gordon Dressler
Reply to  Hans Erren
June 16, 2019 11:47 am

Hans, your mission failure scenario is not correct.

The mission to which you refer was the NASA Mars Climate Orbiter spacecraft. Here is a summary of the correct story (courtesy ):
“NASA lost a $125 million Mars orbiter because a Lockheed Martin engineering team used English units of measurement while the agency’s team used the more conventional metric system for a key spacecraft operation, according to a review finding released Thursday . . .
After a 286-day journey, the probe fired its engine on September 23 (1999) to push itself into orbit.
The engine fired but the spacecraft came within 60 km (36 miles) of the planet — about 100 km closer than planned and about 25 km (15 miles) beneath the level at which the it could function properly, mission members said. The latest findings show that the spacecraft’s propulsion system overheated and was disabled as Climate Orbiter dipped deeply into the atmosphere, JPL spokesman Frank O’Donnell said. That probably stopped the engine from completing its burn, so Climate Orbiter likely plowed through the atmosphere, continued out beyond Mars and now could be orbiting the sun, he said.”

Dodgy Geezer
June 16, 2019 3:08 am

As usual, Arthur C Clarke got there first…

Here, he explains how a spaceship, cut off from Earth communication and with its onboard computers malfunctioning, can nevertheless compute a navigational path back to safety. Indeed, his method will work with no access to electronics whatsoever…

June 16, 2019 5:11 am

Knowing what time it is is HUGE for navigation.

For sailors, knowing how far north or south they are is relatively easy. Knowing your east and west position, on the other hand, requires that you know what time it is fairly precisely. It wasn’t until the invention of a practical chronometer that accurate navigation on the face of the Earth was possible.

In a static universe, you could get your position in outer space by sighting on stars. That’s probably not accurate enough and anyway, the planets are moving. So, the problem becomes much easier if you can measure time accurately.

The other thing that space navigation is similar to is fire control systems in which a naval gunner takes data about his ship’s speed and heading, and the enemy ship’s speed and heading, and the known ballistics of his round, and he tries to calculate where he should point his gun such that the enemy ship and the projectile will arrive at the same place at the same time. So why is the ability to measure time necessary for that calculation? The answer is that the gunner has to figure out the velocity of the ships.

A practical atomic clock for spacecraft is a big deal.

EdA the New Yorker
Reply to  commieBob
June 16, 2019 8:04 am

A dramatization of John Harrison’s work on developing a sea-going chronometer is available at

If they link doesn’t work, just search “Longitude”.

I believe it was an A&E production from 2000. 3 1/2 hours, but very watchable.

Col John S Mosby
June 16, 2019 5:15 am

Now if Tesla can justg manage to launch their Heavy rocket without blowing up and wasting millions and millions of tax dollars. Somehwo Elon Musk always puts liabilities of his companies on the backs of the taxpayer.

Reply to  Col John S Mosby
June 16, 2019 6:29 am

It’s SpaceX, not Tesla and they actually have a great record of successful launches. In fact as poorly as Telsa has been managed, SpaceX is the exact opposite.

They have lost 2 rockets, the last one in 2016 on the pad prior to launch. 79 launches to date and 41 of their stage one cores returned to be reused.

You can bash Musk for lots of reasons, but SpaceX is a very successful enterprise.

Reply to  rbabcock
June 16, 2019 7:22 am

It is successful partly because crony politics

“the Smith provision would require other companies in the race for launch contracts to turn over intellectual property they have developed to SpaceX in order to level the playing field.”

Elon Musk is the best at getting public funds and public help. It doesn’t fit my definition of successful enterprising.

Paul Penrose
Reply to  Javier
June 17, 2019 10:24 am

Those articles are full of disinformation. First off, the “Smith provision” has not been passed, so I don’t know how you can say that SpaceX “has been” successful partly because of something that has not happened yet. Secondly, SpaceX’s competitors, ULA and Boeing have received billions of dollars to do literally nothing. They were paid these monies to just keep them in business so that the US Government had viable launch vehicles. This might make strategic sense for a defense POV, but it seems disingenuous to slam SpaceX for trying to get a cut of that pie when its competitors have already received much more to date. So far, SpaceX has provided valuable services for all the government monies it has received, and at a discount of its competitors. So no subsidies AT ALL.

Paul Penrose
Reply to  Col John S Mosby
June 17, 2019 10:30 am

John Mosby,
What are you smoking? Tesla makes electric cars. SpaceX makes and launches rockets. And for the record, the Falcon Heavy has launched twice successfully, and each one returned two of the three booster cores. The third core on the second launch actually landed successfully on the barge, but then fell over in rough seas before it could be secured. This problem has been resolved for future Falcon Heavy flights. Nobody else can put payloads into space at the cost that SpaceX, partly because nobody else can recover and reuse their boosters, which is a significant fraction of the total cost of a launch. SpaceX is making money and so far has not received any kind of government subsidies. Its competitors can’t say that.

Flight Level
June 16, 2019 5:48 am

Is NASA finally the ultimate public money disposal unit ?

Whatever became from the moon craze ? An unofficial would be Guiness book record for mechanical Hasseblad class cameras, 5771 shots in 4834 minutes of total moon time. One shot every 50 seconds. Ok, roger that, shutter-happy tourists on a leisure trip with nothing else to do.

Then came the climate craze. Wits an absolute record of puzzling data, sea level predictions, doomsday evaluations and finally, yes, the proposal that alien civilizations might weep us if we refuse carbon taxes to save our planet. Roger that too, something completely different, let’s forget about the moon, too many doomsdays to deal with.

Now Mars, the ultimate destination for would be climate refugiees & preferred location for future climate summits.

So what gives ? A consistently negligible utility to dollar ratio at each campaign.

Joe Crawford
June 16, 2019 6:07 am

Clark’s solution sounds similar to what a guy I worked with use to call the first programmable computer he used at the National Bureau of Standards. It was a room full of ladies with mechanical calculators. You provided them with the data arranged in columns and instructions to perform on each column. The only problem was you could only submit the job a couple of times before the ladies recognized it and started complaining. According to him the main advantage of the early digital computers was that they didn’t talk back.

Michael S. Kelly
Reply to  Joe Crawford
June 17, 2019 6:15 am

The Soviets used the same procedure to calculate trajectories for their original space shots. A room full of people (men) would perform the Runge-Kutta integration steps one at a time on a Friden calculator, until they either hit or missed the Moon. If they missed, they adjusted the initial conditions, and did it over again.

The results were engraved on an iridium card, which was read by the rocket’s flight controller as a function of time like a player piano roll. Their rockets were steered into space and then to the Moon by open-loop pitch versus time instructions. The only closed-loop feature was an integrating accelerometer which shut down the engines at the proper velocity.

Joe Crawford
Reply to  Michael S. Kelly
June 17, 2019 7:11 am

Interesting, I didn’t know that. Probably only one step up from that, the IBM flight computer on the Saturn V rocket was a bit-serial computer that executed slightly less than 12 kips.

Carl Friis-Hansen
June 16, 2019 6:30 am

I understand the part about distance and speed, but how does the triangulation work?
My GPS or GLONASS cannot determine my position without three satellites registered.
Maybe some some angular measurement within the signal sent?

Tom Bakewell
June 16, 2019 6:48 am

There are stars that are reasonable radio emitters. Why can’t they be used as beacons in a cosmic GPS setup?

John Dowser
Reply to  Tom Bakewell
June 16, 2019 10:26 pm

Star trackers are already being used but they use (easier to receive) optical frequencies. Even planets could suffice when the tracker has the orbital information ready.

Curious George
June 16, 2019 6:50 am

“being off by even less than a second could mean missing a planet by miles.” By up to 187,000 miles.

Did a settled scientist write this press release?

Sam Pyeatte
Reply to  Curious George
June 16, 2019 7:44 am

That is why we have mid-course correction(s)…and will always have them.

Reply to  Curious George
June 16, 2019 5:31 pm

Last I heard, we don’t have any spacecraft capable of light speed.

Robert of Ottawa
June 16, 2019 7:10 am

Far too bulky. You can already get minituarized atomic clocks; just rad harden them.

You also need good precision in measurement and calculation.

Paul Penrose
Reply to  Robert of Ottawa
June 17, 2019 10:34 am

Not nearly accurate enough.

June 16, 2019 7:23 am

Now to keep it from China’s theft need.

Joel O'Bryan
June 16, 2019 7:48 am

This article isn’t telling the whole story. The SVs that leave Earth orbit to the solar system use highly accurate on-board star trackers to fix their orientation in addition to an IMU. They are not completely dumb. But doppler and ranging from ground tracking stations is currently used to generate precise positions just before and after cruise phase thruster burns for trajectory corrections.
Putting a GPS-like Solar System Navigation System (SSNS) SV at each of of the libration points, aka Lagrange Points, of Earth would allow highly accurate autonomous navigation without use of the ground stations. Then a master ground station would be used to periodically control and update the SSNS SVs. And the robotic space exploring probes would then use the SSNS signals.

The GPS system of SVs all rely on periodic checks, calibrations, error checking and malfunction monitoring, and timing updates from the Master Control Station (MCS) in USAF installation near Colorado Springs, or alternatively at several other ground locations. The SSNS would need such a MCS on Earth to maintain it as well. Then the probes could just use the SSNS signals.

There is also the coming ability of satellites to use known pulsars (fast rotating neutron stars with millisecond rotation periods that send out a microwave pulse due to the co-rotating magnetic field) to accurately navigate. The timing pulses from precisely characterized pulsars in conjunction with an onboard atomic clock would allow precise navigation as well. Such a navigation scheme would need no Earth inputs for control or trajectory corrections and thus would be immune to disruption due to a hostile act.

EdA the New Yorker
June 16, 2019 8:06 am

A dramatization of John Harrison’s work on developing a sea-going chronometer is available at

If they link doesn’t work, just search “Longitude”.

I believe it was an A&E production from 2000. 3 1/2 hours, but very watchable.

June 16, 2019 8:07 am

It bugs me when these articles use units like “one second in 10 million years”. I’m not 100% sure on all these numbers below and a cross check by someone else is welcome. So…the clocks in GPS satellites are stable to parts in 10^12. This is possibly misleading because they are periodically synchronized to ground-based clocks at the US Naval Observatory which are stable to parts in 10^13 or 10^14 (is that right?).

The one second in 10 million years is about 3 parts in 10^14, and that’s backed up by the article linked below. There is other research that suggests the trapped mercury ion technology is capable of stability in the range of parts in to^16, but the compact design apparently doesn’t achieve those levels of stability.

June 16, 2019 10:23 am

One of the problems with navigating in space is having no fixed point from which to calculate position changes. Navigation is tryingt to determine you locations that point when you are moving in 3 vectors. That’s why the Earth stations were critical. Although both rotating and traveling in its orbit they were stable and easily calculated with a stable clock. Both the Space Vehicle and the Earth station has to point their antennas to specific points in 3 dimensional space typically to 1-3 degree accuracy to even find that signal. Once found the equipment can make the minor adjustments for peak reception.

Time is critical foe these calculations. If time drifts on either end they may never point to the needed exact point. This problem is exacerbated with distance. Making synchronizing the clocks critical.

Satellites in Lagrange points are also not stable and take some fuel to make adjustments, both for location and rotation. Rotation being critical for accuracy of pointing that narrow beamed signal when we are talking in terms of planetary distances. Compounding tghe Lagrange locations is the Sun, a radio signal emitter, is always in the background. Your SV receivers and transmitters must find a weak signal or transmit a stronger signal in a heavy white noise background.

Just pointing out some of the probolems,

June 16, 2019 10:40 am

Sounds like a rerun of the longitude problem, which was finally solved by a very accurate clock – the famous Harrison watches – the H4 and H5. See the book Longitude. It is a fascinating story of greed, nepotism, rivalry, and invention.

The H4 watch:
comment image


June 16, 2019 10:43 am

Why do posts seemingly disappear into the aether nowadays….?


Gordon Dressler
Reply to  ralfellis
June 16, 2019 12:24 pm

Can you post an example?

Reply to  Gordon Dressler
June 16, 2019 1:28 pm


Reply to  ralfellis
June 16, 2019 5:37 pm

Why is it that the only posts that don’t disappear, are the posts asking what has happened to all the posts?

Gordon Dressler
Reply to  MarkW
June 17, 2019 2:31 pm


June 16, 2019 10:46 am

“And how can they accurately land on another planet”

To land on another planet one needs to know where one is in relation to that planet; not ones location relative to anything else. So RADAR, LIDAR or the like mounted on the space craft would be used to measure the relative position of the planet and that would get more accurate as the space craft got closer.

Gordon Dressler
June 16, 2019 11:56 am

Interesting and true: The Earth’s constellation of GPS spacecraft use 212 quasars in the universe as their foundation for accurate position determination. See

June 16, 2019 2:53 pm

There’s No GPS in Deep Space

That’s not exactly true. There are pulsars all over the galaxy. The Pioneer plaque uses 14 pulsars to pinpoint our location in the galaxy. Since the periods of pulsars change at a known rate, the launch time of the Pioneer craft can be determined too.


Gordon Dressler
Reply to  Jim Masterson
June 16, 2019 7:17 pm

I think that first statement was a simple observation that GPS, as based on Earth-orbiting satellites, was not available/usable in deep space. I don’t think the statement’s author meant to imply it was impossible to develop the EQUIVALENT of a GPS system that could operate in deep space.

June 16, 2019 6:58 pm

Pure hype. You do not need clock to navigate using triangulation and some type of GPS system whether on Earth orbit or on Earth or on Moon. You need coordinates of GPS stations at times of signal emission by GPS stations and the arrival times of signals that were sent by them semi-simultaneously. The times of emission are the timestamps in the signals from the GPS stations and the GPS stations’ positions are calculated from ephemeris for each GPS station that is being updated every several days and sent to the receiver. Knowing time differences between times sent the position of the receiver can be calculated. In other words the clock of the receiver does not need to be synchronized with clocks of GPS stations. However you need four GPS stations not just three.

John Dowser
June 16, 2019 10:23 pm

“spacecraft flying beyond Earth’s orbit don’t have a GPS to find their way through space”

Some optical, automated navigation or positioning is done by star tracker devices. But the problem of accuracy and relative timing remains of course. The star tracker makes some basic navigational actions without transmissions from Earth or relay stations possible, although they are complex devices.

%d bloggers like this:
Verified by MonsterInsights