Where did the moons of Mars come from? That’s a question scientists still can’t answer. We know that Earth’s moon was likely formed from a giant impact on our planet about 4.5 billion years ago. Some moons in the solar system, such as several of Jupiter’s smaller satellites, appear to be captured asteroids. It remains unclear which of these two formation routes holds true for Mars’s moons, Phobos and Deimos—but we may soon have an answer. A Japanese spacecraft launching next year will attempt to bring samples back from Phobos. The mission will build on exciting new results from a United Arab Emirates (U.A.E.) orbiter at Mars that suggest a planetary origin for the two moons. “There’s room to be surprised, but I think we’re going to figure it out,” says Jemma Davidson of Arizona State University.

On April 24 the U.A.E. announced that its orbiter, Hope, had studied the smaller of Mars’s two moons, Deimos. The spacecraft returned some of the best data and images of Deimos yet from as low as 100 kilometers above the moon’s surface. Those results suggest the Deimos’s composition more closely matches Mars than that of a class of asteroids that was previously flagged as the likely raw material for Deimos and Phobos alike: D-type asteroids in the outer asteroid belt between Mars and Jupiter. “We don’t believe that [Deimos] is an asteroid,” says Hessa Al Matroushi, science lead of the mission at the Mohammed Bin Rashid Space Center in Dubai.

To find out for sure, scientists want to return samples of Phobos to Earth. An attempt by Russia to do so ended in failure in 2012, when its Phobos-Grunt spacecraft crashed into the Pacific Ocean shortly after launch. “It never got out of Earth orbit,” says John Logsdon, a space historian and professor emeritus at George Washington University’s Space Policy Institute. The Japan Aerospace Exploration Agency (JAXA) is hoping to avoid the same fate with its Martian Moons eXploration (MMX) mission. The solar-powered spacecraft, expected to launch in September 2024, weighs in at more than three metric tons and is roughly the size of an SUV. It will aim to enter Martian orbit in August 2025 before sidling up to Phobos in 2026 to scoop samples and return them to Earth by 2029. The mission is “super complex” but should be highly rewarding, says Patrick Michel of the Côte d’Azur Observatory in France, a European collaborator on MMX and a member of the mission’s science board.

On April 17 NASA and JAXA announced they would be partnering on the mission. As part of the partnership, NASA selected 10 U.S. scientists to work on MMX and will also supply two instruments for the spacecraft. “We’ve got great partners at JAXA, and they are leading this ambitious mission to bring back the first samples of the Martian moon Phobos,” said Bill Nelson, NASA’s administrator, in a video message posted to Twitter. “Together, we’re going to deepen our knowledge of the solar system.”

Of Mars’s two moons, Phobos is slightly larger. Both are irregularly shaped, like potatoes. Phobos is about 27 km across on its longest side, and Deimos is 15 km across. Phobos is also the closer of the two to Mars. It orbits just 6,000 km above the surface and completes an orbit every seven hours and 39 minutes. Deimos, at more than 23,000 km in altitude, takes slightly more than 30 hours to orbit. Both moons have been imaged by several spacecraft before, most notably by NASA’s Viking 2 orbiter in 1977 and by the Mars Reconnaissance Orbiter in the 2000s and even by the Curiosity rover from the surface of Mars in 2013. But no spacecraft has ever landed on either moon.

Japan’s MMX mission will attempt to change that. It builds on the success of the nation’s asteroid sampling missions, Hayabusa and Hayabusa2, which returned samples of asteroids in 2010 and 2020, respectively. Both of those, however, spent mere seconds brushing across the surfaces of their targets. MMX will land on Phobos in two locations and spend two hours on the surface collecting about 10 grams of material in total. “That’s a big difference with Hayabusa,” Michel says. Surface operations on Phobos pose many challenges because the moon has just a thousandth of Earth’s gravity—and an uneven gravity field at that, given its unusual shape. MMX will gather samples using two methods: a coring sampler on an extendable arm to collect specimens from deeper than two centimeters and a pneumatic sampler to kick up material from the surface.

Before MMX collects its samples, however, it will seek to ensure a smaller landing takes place. In 2026 or 2027 the spacecraft will deploy a small rover on the surface, developed by scientists in France and Germany. The rover, the size of a microwave, will be dropped from a height of 45 meters when the spacecraft performs a practice landing attempt. After tumbling on the surface, the rover will then be righted by its four extendable wheels to begin a 100-day mission. The moon’s weak, irregular gravitational pull means that the rover, despite weighing just 25 kilograms, will not be able to travel faster than a snail’s pace because it would otherwise risk launching itself into space.

“If we’re going quicker than 80 millimeters per second, we might flip over the rover or even leave the Phobos system,” says Markus Grebenstein of the German Aerospace Center, the project manager for the rover. Accounting for the rover’s limited lifetime, that speed limit “basically restricts our range to about 100 meters.” Even so, the rover should prove invaluable. It will study the surface of Phobos and give the main MMX spacecraft vital information on the moon’s surface properties that will be incorporated into the two landing attempts. The rover will also test robotic operations on a small body such as Phobos. A stretch goal might be to push the rover to its limits by spinning up its back wheels at the end of the mission in an attempt to flip it. “The rover would easily be able to do a backflip,” Grebenstein says. “We might be allowed to do experiments like that at end of its life.”

The target for MMX will be sampling “the most pristine material on Phobos,” Michel says, which may include hints to its origin. The samples may have a hidden bonus, too. The surface of Phobos is thought to be covered in some material that was ejected from Mars via impacts and then settled on the moon. So when Japan brings its samples to Earth in 2029, they may well contain the first pristine ones collected from the planet itself, beating NASA’s multi-billion-dollar Mars Sample Return effort, which is not expected to send samples back to our planet until 2033 at the earliest, by a considerable margin. MMX’s samples are unlikely to contain any evidence of past life or habitability on Mars, but they may provide useful information about its past geology. “We hope we can capture them in the sampling mechanism,” Michel says. “We could have the first retrieved samples from Mars with this mission.”

After its two landings, MMX will leave the surface and send its collected samples back to Earth in a capsule. While the main spacecraft itself will stay in Mars orbit, subsequently performing flybys of Deimos to study that moon from afar, the sample capsule will touch down in an Australian desert in July 2029. Davidson is one of the scientists selected by NASA who will then investigate its samples back on Earth. “By looking at the minerals, we’ll be able to tell if it’s a mineral from Mars or a captured asteroid,” she says.

If the samples prove to be captured asteroids, this finding will pose interesting implications for how they migrated from the outer asteroid belt to Mars. But if they are pieces of Mars, formed by an impact early in its history, that poses its own problems—not least by raising the question of how smaller objects such as these formed around a planet, compared with the size of our own moon around Earth, which is unfathomably larger at some 3,500 km across. “It doesn’t fit the models we have for what material from a giant impact would look like,” Davidson says. “Whatever we figure out, we have to rethink what we’ve assumed we know about these processes.”

MMX and Hope represent a renewed interest in the moons of Mars, which were suggested by the Planetary Society in 2015 as prime locations to begin human exploration of the Red Planet. “If we couldn’t send humans to the surface of Mars, maybe we could send them to rendezvous with Phobos and Deimos,” says Logsdon, a co-author on the Planetary Society report. Now we are closer than ever to working out where they came from, which could help us understand more about how our solar system and its myriad of planets, moons and asteroids came to be. “Understanding how the moons formed is really fundamental to us understanding the dynamics of our solar system,” Davidson says.