Everyone loves a story about a small but daring character that overcomes all obstacles to become a hero at just the right moment. This week saw one such story play out in dramatic detail—and it was set in space.
On November 27, and the world’s attention is fixated on Mars. A lander would soon enter the Martian atmosphere and land on its surface, collecting troves of new data .
Earlier landings have been vexed by a lack of steady communications due to the movement of the satellites around Mars. Without those sats to serve as data relays, it could be hours for news about the lander’s condition to return to Earth.
But on Monday, there are two solutions flying past Mars. A pair of experimental satellites—each the size of a briefcase—were on hand to capture and send near real-time data as the craft descends to its new Martian home. Their official names are Mars Cube One-A and Mars Cube One-B, but everyone calls them “MarCO.”
As the lander streaks through Mars' thin atmosphere, these diminutive twins are going to become the center of attention as information becomes incredibly precious. All eyes turn to the MarCO-A and -B for confirmation that the landing is going well, hoping they can deliver in minutes what would usually take hours.
But in order for this intrepid duo to rewrite the rules on planetary landings, the spacecraft had to get to Mars first—a feat easier said than done.
A CubeSat Origin Story
MarCO-A and MarCO-B were born to be “secondary payloads,” which is good and bad for a spacecraft. While the two sats secured a one-way ticket to space, they remained subordinate to the main mission. In this case, the InSight lander, and in an absolute worst-case scenario, the two twin satellites are expendable.
CubeSats have born-to-die appeal. In the case of MarCO, the engineers at JPL constructed them from standardized, commercial CubeSat tech, making them much less expensive. In comparison, MarCO cost $18.5 million while InSight tallies at around $823 million.
CubeSats by their nature weigh less than larger ones, and so they are also cheaper to launch and flagship projects have staff measured in hundreds, whereas MarCO could be measured by the handful. The lack of expenses make CubeSats ideal for daring missions, and constellations of them may become ubiquitous in low-Earth orbit.
But first they need to prove they are tough enough and smart enough to handle deep space.
Ready for Launch
The MarCO twins’ first challenge is a six-month trip to Mars—taken on their own. It would make sense to have the CubeSats cling to the InSight lander like baby possum during the trip. But a Class A mission like the Martian lander isn't about to let a couple of experimental satellites ride along the whole way.
MarCO JPL engineer Brian Clement tells Seniorhelpline that the protective stance toward the main mission meant the team had to plan on detaching the sats just after launch. “The mission architecture of a secondary is designed first and foremost to do no harm to the primary payload,” Clement says.
On May 5, 2018, InSight and its two wingmen launch atop an Atlas V from Vandenberg Air Force Base. “We didn’t even ride in the payload fairing,” says JPL navigation engineer Brian Young, one of two navigators for the MarCO mission. “They were just tucked at the bottom of the second stage.”
The MarCO sats become pioneers days after ejecting from the rocket. As CubeSats on a deep space mission, they become the first to perform maneuvers in deep space. There are four trajectory correction maneuvers planned for the six-month journey, using the spacecraft’s thrusters to steer toward Mars and a set of reaction wheels that point the spacecraft in the right direction, using the torque to rotate the craft around its center of mass. This is an easy way for a spacecraft to move around without using engines.
The MarCO team used standard astronavigation software to plot the course, but the JPL navigation engineers created new tools to streamline the process. A secondary payload doesn’t get a dedicated team; the JPL staff working the twin sats are expected to only devote between a quarter of their time to the project. Even during peak operations that percentage rises to just half.
“I hope some of what we learned will flow into the bigger missions,” Young says. “We tried to create tools that cover all the basics and they can be combined for whatever a mission needs.”
But just as the MarCO team prepares to start the sats' historic deep space maneuvering, the sensors alert them to some bad news. MarCO-B has a problem.
The Deep Space Navigators
Class A missions are staffed with between five and 12 JPL navigators, but MarCO only had two: Brian Young and Tomas Martin-Mur. The pair’s skills are about to be put to the test.
Spacecraft make course changes called Trajectory Correction Maneuvers (TCMs), which are usually done with thruster bursts that are timed to the second. “We were doing the first test for a TCM,” Young says. “We were doing this to calibrate the engines. We did a small maneuver in directly along the Earth line.”
The engineers’ idea is to fire the thrusters on a path away from Earth so that Doppler radar can measure the velocity change. This data will be used to predict future TCM burns.
But something is wrong. Pressure sensors first indicate that there is fuel leaking into MarCO-B’s plenum, a pressurized chamber that holds the gas before valves route it through the thrusters. Some thrusters are at an angle, to help its orientation (attitude control), while others are straight to provide straight-on push.
Dumping that gas didn’t help as the pressure sensors shoot up again. “It kept building up,” Young says. Even worse, the tank leak into the plenum isn't the only one. One of the angled thrusters is leaking fuel into space, too. Now the leak is a real threat, since it’s producing enough unwanted thrust to create trajectory changes.
The leak is not constant, and while sometimes the release is only two meters per second, other unwanted thrust climbs as high as five meters per second, which is about a quarter the size of one of the four planned TCMs.
The small MarCO team gather immediately to discuss the issue. Groups of ten and then 20 determine the scope of the problem and then start figuring out a way to pilot the craft despite its infirmity. “That was a weekend spent trying to figure out what was going on and how to bring the spacecraft back into a stable state,” Young says.
Managing the leak means applying the hardware that steers the craft. The reaction wheels can spin to absorb and counteract the angular momentum caused by the gas coming from the thruster. However, the reaction wheels may build up too much momentum that it exceeds the maximum speed of the wheel.
“We’ve got this thrusting that we can’t turn off,” Young says. “So what can we do with it?”
By playing off this interaction between spacecraft controls, the two-person navigation team regain control of where the sat is going. But to keep it under control means constant vigilance.
“Instead of just letting it leak, every hour or half hour we would then open all the valves and release the gas in that plenum so that we would get thrust in a controlled direction,” Young says. “We could control all the valves but one that didn’t close all the way. So instead of letting it all leak out the one we couldn’t fully close, we’d open the ones that wouldn’t cause any changes in angular momentum, just a velocity shift in the right direction.”
The pair also orient the spacecraft to take advantages of the pressure generated by solar radiation, which can counteract some of the unwanted motion, Young says.
As the navigators get more practice, they were able to predict the changes and were better prepared to correct them.
“Doing the maneuvers was always a little bit risky,” Young says. “There’s always the risk that another bigger leak would occur and cause a deviation of the trajectory that you can’t correct.”
The lessons of the flight, and the hourly attention given to it by the two devoted navigators, begin to apply to MarCO-A.
“Because of what was happening with B, we never wanted to open the valve to do any desaturation burns, when the internal wheels spin up too fast and you have to fire the thrusters to remove that angular momentum. So we found attitudes that would de-spin the spacecraft slowly,” Young says.
Things appear to be under control, but as Mars gets closer, NASA issues press releases that take a measured but almost negative tone regarding the sats’ future: “Should the CubeSats make it all the way to Mars, they will attempt to relay data to Earth about InSight's landing. InSight won't rely on either CubeSat for that data relay, however; that job will fall to NASA's Mars Reconnaissance Orbiter.”
The expectations have been set low for MarCO-A and -B even before launch. But the pair of sats were going to prove what they could do when it counted the most.
MarCO at Mars
Landing day is the Super Bowl at the Jet Propulsion Lab. News trucks flock, satellite antennas raise to the sky. Reporters crowd all available seats at designated newsrooms, auditoriums, and overflow trailers. A pop-up JPL store sells out of their supply of InSight sweatshirts, coffee mugs, and other paraphernalia. There are no MarCO T-shirts for sale.
For all the talk about the science, which will take years to gather and decades to fully decipher, the engineering feat of what NASA calls entry, descent, and landing (or EDL) on Mars takes a bulk of the mission’s attention. It’s one of the few times there will be a clearly defined moment, a peak of celebration and achievement—or disappointment.
So when the secondary payloads—MarCO-A and -B—steal the show during EDL, it happens when the world’s attention is fixed on the Red Planet.
Both MarCO sats are essentially built to capture the UHF signal from the lander as it’s descending and rebroadcast it on an X-band frequency toward Earth. There are two CubeSats because the shape of the InSight antenna leaves a dead zone that might hamper the communication between just one, especially if the lander is thrashing around during the entry. MarCO-B parks in the northern hemisphere and A covers the south.
Without these two sats, the team of Earth must wait for hours to get trickles of data and images from the surface. It’s not mandatory for the mission to succeed, but the rapid data delivery will certainly make the moment even sweeter by confirming touchdown in near real-time.
The mission control announcements begin as InSight enters the thin Martian atmosphere. First, MarCO-A establishes , and MarCO-B soon follows. It’s the first applause of the day, an early win that shows that the essential function of the spacecraft is working. The lander is not moving around as much as feared, so both sats connect without encountering the antenna’s null zone.
Now it’s time for the CubeSats to deliver. The sats shuttle telemetry signals from the InSight lander as it first uses a heat shield to slow down, then opens a supersonic parachute. The mission control rooms in Pasadena and Denver are getting reports as they happen—plus the nearly seven-minute delay it takes for a signal to reach Earth from Mars. There are no disruptions, lost data, or frame-rate drops.
When the call quickly comes “touchdown confirmed,” the moment of frenzied dancing, hugging, and hand slapping is courtesy of MarCO. Within about 15 minutes after confirmation, the two sats give one more gift—the first image of the probe’s home on Elysium Planitia.
It’s taken from behind a smeared protective cap, but the media shares the image worldwide immediately. “This image is actually a really good argument for why you put a dust cover on a camera,” says Bruce Banerdt, InSight’s principal investigator.
The InSight team devours each detail—the camera is showing less than two degrees of tilt, which tells them the lander is sitting nice and flat for its science. The photo’s foreground also shows a rock with some dust sculpted around its base. That seems to indicate sandy soil, a good sign for future drilling experiments.
“That first image showed us how difficult it would be to place the instruments, ” says science system engineer Elizabeth Barrett. “Luckily it looks like nothing is in the way.”
Just after the landing, MarCO-B captures an image of Mars from about 4,700 miles away. The arc of MarCO’s antenna stretches to one side, and the curve of Mars firmly centered. Although MarCO-B is only supposed to be a communication relay, it’s also proven to be a capable investigative tool. “We are able to do some science there as well,” says MarCO team lead Andrew Klesh, saying that atmospheric scientists have already started examining the image.
The image appears to the public in a packed auditorium at JPL, crowded with media, visitors, and as many InSight team members as can make it. There is a collective gasp and applause as it appears on screen.
“This image is really our farewell to InSight, our wish for good luck,” Klesh says. “And our farewell to Mars itself.”
Into the Sunset
There’s no great sendoff for the two MarCO satellites. For two weeks they will collect and send data related to their operation during EDL. After that their official role will end. The two will continue in separate elliptical orbits around the sun, embarking on an eternity of silent spaceflight. “They basically will be going opposite directions from the sun,” Young says.
Their legacy will not fade fast. “We’ve shown these kind of craft can support missions like this,” says Klesh.
That’s inarguably true, especially since not every destination in the solar system will have numerous satellites around them like Mars does. The idea of sending CubeSats with missions is being taken seriously within the industry as well as NASA.
Some of the ideas for future missions feature mothership concepts, where a main spacecraft leaves Earth with a bunch of CubeSats attached, says Tim Linn, a senior space engineer at Lockheed Martin and InSight’s deputy program manager. “They all get deployed once you get to your destination.”
Another industry that is eyeing the mission are CubeSat launchers. Any use for this class of spacecraft means more revenue potential for sat launchers like Virgin Galactic.
“We can launch future CubeSat probes on dedicated small launchers, equip them with ion engines to travel ultra-low energy trajectories, enabling exploration of thousands of solar system objects,” George Whitesides, CEO of Virgin Galactic and The Spaceship Company. “The future is exciting!”
MarCO-B may have lost its twin and InSight lander, but it’s not entirely alone. JPL staff will monitor them to determine how long the sats can endure, using whatever time engineers can scrounge from the Deep Space Network.
For Young, his relationship with MarCO is also not over. As the craft moves off into space, the JPL space navigator will continue to correct for the fuel leak. “We have the leak under control and we know how to deal with it,” he says. “So it can go a long time. It really depends on the life of the electronics.”
So the careful tending of these brave CubeSat explorers will continue—at least for now.
“We’re still in negotiations to see what we’ll do with that, since it takes resources and money," Young says. "But…it will be interesting to see how long they’ll last.”