Japan's MMX Lifts Off October 19 to Steal a Piece of Mars' Moon

In seven days, a rocket that has already failed twice will attempt something no spacecraft has ever done: steal a piece of Mars' moon and bring it home. The Japan Aerospace Exploration Agency's Martian Moons eXploration spacecraft, better known as MMX, is scheduled to lift off on Monday, October 19, at 3:41 p.m. ET from the Tanegashima Space Center in Japan, riding an H3-24L rocket toward a rendezvous with Phobos, the larger of Mars' two small moons.

The plan is audacious in scope and patience. MMX will slip into orbit around Mars in August 2027, survey Phobos for years, descend to scoop up at least 10 grams of soil, and deliver it to Earth in 2031. At 4,480 kilograms (9,880 pounds), it is the heaviest planetary payload Japan's H3 rocket has ever carried. If it works, scientists will hold the first physical matter ever returned from the Martian system. NASA previews its contributions in a media briefing on Thursday, October 15, at 4 p.m. ET.

The fifty-year mystery ten grams could solve

Phobos and its smaller sibling Deimos have puzzled planetary scientists since the 1970s. The captured-asteroid camp argued the moons were ancient rocky bodies snared by Mars' gravity billions of years ago. Their dark, carbon-rich appearance supports that, but their nearly circular orbits hugging Mars' equatorial plane do not: random captures should trace wild, tilted paths. The giant-impact camp argued a huge body slammed into early Mars, blasting a debris ring into orbit that clumped into two moons, like Earth's Moon. That explains the tidy orbits, but the moons' dark spectral signature looks nothing like the bright basaltic ejecta such an impact should produce.

In 2025 the debate got more complicated. A paper in Icarus by researchers at Durham University and NASA's Ames Research Center added a third hypothesis: disruptive partial capture. An asteroid that wandered too close to Mars was torn apart by tidal forces into fragments, many captured into a ring that accumulated into Phobos and Deimos. Simulations show this can produce the moons' circular, equatorial orbits, and the resulting bodies would carry a composition between pure asteroid material and Martian ejecta. A 2026 Nature Astronomy study added another wrinkle, reporting that Deimos was radically reshaped by an asteroid impact, based on images ESA's Hera mission captured during a March 2025 flyby.

Orbital observations cannot settle this. Only a physical sample can reveal the isotopic ratios, the chemical fingerprints that definitively tie a rock to a parent body. That is what MMX is going to get: the first tool capable of testing all three theories at once.

Ten grams of moon dust is about to settle an argument that telescopes could not resolve in fifty years.

Landing on a moon that barely pulls back

Phobos is an orbital mechanics puzzle no previous JAXA mission has faced. The moon is roughly 27 kilometers (16.8 miles) at its widest and 18.2 kilometers at its narrowest, and its surface gravity is about 1/2,000th of Earth's: enough to hold a landed spacecraft, but far too weak for a conventional satellite orbit. A spacecraft approaching at normal low-orbit altitude would be pulled away by Mars' gravity rather than anchoring around the moon. MMX solves this with quasi-satellite orbits, trajectories that loop around Mars while staying near Phobos, flying a series of progressively tighter ones to map the moon from decreasing altitudes.

The touchdown will be compressed into about 2.5 hours, because Phobos whips around Mars every 7.65 hours and mission designers effectively get one pass per orbit. To make those hours count, the spacecraft carries two independent sampling systems, either of which alone can meet the requirement. The coring sampler drives into the regolith to collect material from deeper than 2 centimeters (0.8 inches) below the surface, shielded from solar radiation and micrometeorite gardening. The pneumatic sampler, built by Honeybee Robotics, a Blue Origin company, fires pressurized gas to kick loose particles into a container. MMX may attempt two touchdowns, depending on the first.

The scout that falls from the sky

Before the main spacecraft lands, a smaller traveler gets there first. IDEFIX, a 25-kilogram (55-pound) wheeled rover named after the dog in the French Asterix comics, was jointly developed over nearly a decade by France's CNES and Germany's DLR. No rover has ever operated on a moon of Mars. The mothership will release it from about 40 meters (131 feet) above the surface, and it will free-fall in near-zero gravity until it hits the ground, then right itself entirely autonomously before up to 100 days of exploration with cameras, an infrared radiometer, and a Raman spectrometer.

The rover lands first for a practical reason: its ground-truth observations of soil texture and terrain, relayed through the orbiting mothership, will help JAXA pick a safe touchdown zone for the much heavier main spacecraft. Control is shared between Toulouse and Cologne.

A genuinely international spacecraft

JAXA calls MMX the largest international collaboration it has ever led. NASA contributes MEGANE, a gamma-ray and neutron spectrometer built by Johns Hopkins APL and Lawrence Livermore, plus Deep Space Network tracking and funding for American scientists, all backed by the Discovery Program. France's CNES co-leads the IDEFIX rover and contributed the MIRS infrared spectrometer. ESA provides deep-space tracking through its Estrack network. And Japan's NHK installed an 8K/4K Super Hi-Vision camera that will capture the highest-resolution video ever taken from the Martian system, with the data stored in the return module so it physically travels back to Earth inside the sample capsule.

The rocket has something to prove

MMX rides the H3-24L, the most powerful configuration of Japan's next-generation heavy-lift rocket: two liquid LE-9 core engines, four solid strap-on boosters, and a long fairing. It is also the first time the H3 has been asked to inject a payload onto a deep-space trajectory, and its road here has been rough. The debut in March 2023 failed when the second-stage engine did not ignite, destroying the payload; a December 2025 flight failed as well. On June 12, 2026, H3 Flight No. 6 returned to flight in a lower-cost configuration, restoring confidence. MMX flies as Flight No. 10. For a rocket rebuilding its reputation, carrying Japan's most ambitious deep-space mission is both a vote of confidence and a very public test.

What happens next

The launch window is unforgiving. Earth and Mars align for efficient transfers only about every 26 months, and JAXA holds a backup window through November 7. Miss that, and the mission waits roughly two years. On schedule, MMX reaches Mars in August 2027, surveys both moons (studying Deimos by flyby only), deploys IDEFIX in late 2028 or early 2029, collects its sample, departs in 2030, and re-enters Earth's atmosphere in 2031 over the Woomera desert in South Australia, the same zone used to recover Hayabusa2 in December 2020. That mission returned 5.4 grams from the asteroid Ryugu, material later found to contain amino acids and water-bearing minerals.

China's Tianwen-3 is targeting a return of Martian surface soil around 2031, potentially arriving at Earth around the same time as MMX's Phobos capsule. Together they would make the early 2030s the most productive decade of Martian exploration, and it all starts with a launch next Monday.