A rocket that lands itself, and a Venus mission saved from a five-year miss

01 What happened

The RV-X flew from JAXA’s Noshiro Rocket Testing Center in Akita Prefecture, on the north coast of Japan’s main island, on 11 July 2026.[2][3] The vehicle stands 7.3 metres tall and 1.8 metres across, and it rests on four shock-absorbing landing legs.[2][5] Under a single hydrogen-fuelled engine it rose to about 11 metres, translated roughly 15 metres horizontally while keeping its body vertical, and descended to a controlled landing on the far side of the pad, the whole flight lasting under a minute.[2][1]

This was the program’s first successful flight, and it followed a false start. An attempt in March 2026 was called off after a malfunction in the quick-disconnect, the coupling that feeds propellant to the vehicle and detaches at liftoff.[3] The engine that flew had by then accumulated 165 firings on the ground, a tally meant to prove that one engine can be lit, shut down, and lit again without being discarded.[2][3] “We completed the test flight properly and obtained data that we had wanted,” said Takashi Ito, a JAXA research and development manager.[2]

The altitude is deliberately small; the demonstration is about procedure, not height.[4] JAXA describes the goal as verifying vehicle operation, maintenance, and pad setup, and maturing the navigation, guidance, and landing-control logic that a vertical-takeoff-and-vertical-landing craft needs and an expendable rocket does not.[4] The horizontal move matters because a returning stage must correct where it comes down, not simply fall back onto its launch point.[1] Reported figures differ slightly by source, with the altitude given as about 10 to 11 metres and the sideways distance as 15 to 16 metres; they are reported here unaveraged.[1][2][5]

02 The physics of why it is hard

The engine burns liquid hydrogen, the lightest chemical fuel and the one that yields the highest specific impulse, the thrust produced per unit of propellant flow, measured in seconds.[2] Hydrogen buys performance at a steep handling cost, because it must be kept liquid near 20 kelvin, roughly minus 253 degrees Celsius, and it is so light that tanks must be large and heavily insulated to limit boil-off, the slow warming that turns liquid back to gas.[3] A hydrogen machine that must survive repeated ignitions, rather than fly once, is a harder engineering problem than an expendable one, which is why the flight engine was fired 165 times first.[2]

The landing itself is a balance problem. A nearly empty vehicle weighs little, so the engine must throttle down far enough to hold the thrust-to-weight ratio, the ratio of engine push to vehicle weight, close to one, or the craft accelerates upward instead of settling gently.[4] Holding a tall, narrow body vertical while it also moves sideways demands closed-loop guidance, meaning the vehicle senses its own position and attitude and corrects the engine’s aim many times a second, a control task Japan’s single-use rockets never had to solve.[4]

The deeper aim is operational rather than aerodynamic. Reuse only pays if a recovered vehicle can be inspected, serviced, and readied on the pad quickly and cheaply, so the test rehearsed maintenance, handling, and pad setup as much as flight.[4] An 11-metre hop exposes the ground procedures and the turnaround work that decide whether reuse actually lowers cost, which is the part a purely higher or faster flight would not reveal.[1]

03 What to watch

The next concrete milestone is a higher flight. JAXA plans to fly RV-X to about 100 metres, roughly ten times this altitude, adding hover and lateral maneuvers before landing, though the agency says it will first review the data and decide whether to reuse the same engine.[4][3] Beyond RV-X sits CALLISTO, a larger vertical-landing demonstrator that JAXA is developing with France’s CNES and Germany’s DLR, with Mitsubishi Heavy Industries as an industrial partner, aimed at the toss-back and recovery of an actual launcher stage.[2][5][4]

01 The story

Venus turns on its axis once every 243 Earth days, yet its dense cloud deck races around the planet in about four days. This mismatch is called super‑rotation, an atmosphere circling roughly sixty times faster than the ground beneath it, and no one could say what keeps it spinning. Akatsuki, also named Planet‑C, was built to watch that weather from orbit with a suite of cameras spanning the ultraviolet to the mid‑infrared. It launched on an H‑IIA rocket on 20 May 2010 at a mass of 517.6 kilograms, part of a program that cost roughly ¥14.6 billion for the satellite and ¥9.8 billion for the launch.[1][5]

The plan was a single decisive burn. On 7 December 2010 the 500‑newton bipropellant main engine, the orbital maneuvering engine, lit to slow the craft into a tight loop around Venus with a 30‑hour period and an apoapsis, the orbit’s farthest point, near 80,000 kilometers. It shut down at 158 seconds instead of the intended 720. The spacecraft passed behind Venus, tumbled, and emerged not in orbit but back in orbit around the Sun.[4][1]

The failure review traced the cause to a check valve on the fuel line. A solid salt had formed inside it, generated when fuel and oxidizer vapors mixed with the helium used to pressurize the tanks. The valve throttled the fuel while oxidizer kept flowing, so the mixture ran steadily leaner and the combustion temperature climbed until the ceramic throat of the nozzle cracked under thermal stress. The main engine was effectively destroyed. In 2011 controllers took a hard step and dumped the spacecraft’s dinitrogen tetroxide oxidizer, now dead weight, to lighten the craft for whatever came next.[4][6]

Getting a second shot meant waiting for the geometry to repeat. Akatsuki was placed in a resonant orbit that made nine loops around the Sun of about 203 days each while Venus made eight of 225 days, bringing the two back together after five years. Through that wait the craft flew closer to the Sun than it was designed for, near 0.6 astronomical units instead of 0.7, and absorbed a heat load roughly a third higher than intended. On 7 December 2015 it fired four 23‑newton attitude‑control thrusters for 1,228 seconds and slid into a long ellipse of about 400 by 440,000 kilometers with a period of 13 days and 14 hours. A later trim pulled the period down to about 10 days and 12 hours.[6][2][3]

02 The hard part

Reaction‑control thrusters are meant to nudge a spacecraft’s orientation, not to perform an orbit‑insertion braking burn. Four of them together produced 92 newtons, less than one‑fifth of the 500‑newton main engine the mission was designed around. With that little thrust there was no reaching the planned 30‑hour orbit; the best achievable was a far larger, slower ellipse. The team had to conserve enough hydrazine across five years to make the attempt at all, having already flushed away the oxidizer that the dead main engine could no longer use.[6][4]

The waiting itself was the other adversary. Electronics and propulsion had to survive well past a 4.5‑year design life while running hotter than specified on the inner leg of the resonant orbit. A monopropellant burn of 1,228 seconds is long and unforgiving; a premature shutdown, as in 2010, would have ended the mission for good with no third pass on the calendar. That the craft threaded all of this to a stable orbit, if not the intended one, is the measure of the engineering.[6][2]

03 Why it mattered — measurably

Before Akatsuki, Japan had never placed a spacecraft in orbit around another planet. Its only prior attempt, the Mars probe Nozomi, suffered a propulsion failure and sailed helplessly past Mars in 2003. The 2015 insertion changed that record, making Akatsuki Japan’s first working planetary orbiter and, by the time it fell silent, the last active spacecraft anywhere at Venus.[6][8]

The salvage also converted a lost mission into a scientific one. Cloud‑tracking from the recovered orbit let researchers measure how angular momentum moves through the atmosphere, and in a 2020 paper in Science the team showed that thermal tides, driven by solar heating, do the work of accelerating the roughly 100‑meter‑per‑second super‑rotation at low latitudes, while other waves and turbulence act against it nearer the poles. Earlier, in a 2017 paper, researchers reported a bow‑shaped feature the craft had imaged, stretching nearly pole to pole and holding still over a highland region for days, read as a stationary gravity wave rising from the surface below.[7][5]

The honest accounting notes what was given up. The intended 30‑hour orbit was never reached; the achieved period of ten to fourteen days gave coarser time sampling of a fast‑moving atmosphere, and the two near‑infrared cameras failed in 2016. Against that, the mission ran until contact was lost in 2024 and was formally closed in 2025, roughly fourteen years after launch and about triple its 4.5‑year design life. Whether a clean 2010 insertion would have returned more remains unknowable; what the record shows is a mission recovered from zero to a decade of Venus data.[1][8]

04 Echoes today

Akatsuki’s fingerprints show up in two places. Its super‑rotation measurements are now reference data for the atmospheric models behind the next wave of Venus missions, which return only later this decade after Akatsuki left the planet without a working orbiter. And its recovery stands as a worked example that a crippled spacecraft is not automatically a dead one, that propellant discipline, patience across years of coasting, and a willingness to insert on thrusters never meant for the job can pull a mission back from a clean miss.[7][8]