Moon Landing Attempts: 2023 to 2025

Introduction

The renewed global push to explore the Moon, particularly its resource-rich south pole, has reignited interest and investment in lunar missions. This new era is defined not only by the ambitions of established national space agencies but also by the emergence of a dynamic commercial sector. However, a series of high-profile successes and dramatic failures between 2023 and 2025 has starkly illustrated that landing on the Moon remains one of the most difficult undertakings in space exploration. These recent attempts are more than just individual missions; they represent a collective, high-stakes stress test of different technologies and programmatic philosophies. The outcomes are generating a rapid, publicly visible feedback loop that is accelerating the learning curve for the entire industry, shaping the future of humanity’s presence on the Moon.

The Unforgiving Gauntlet: Why Landing on the Moon is So Hard

Decades after the Apollo program, touching down softly on the lunar surface continues to be a formidable engineering challenge. The difficulty lies not in a single obstacle but in a series of interconnected problems that must be solved flawlessly by an autonomous machine, millions of miles from human help. The recent missions have shown that failure at any stage of this final gauntlet can be catastrophic. The core challenge is one of extreme system integration, where navigation, propulsion, and hazard avoidance must work in perfect harmony under immense time pressure, with zero margin for error and no possibility of direct human intervention.

Navigating Without a Net

A spacecraft approaching the Moon cannot simply look up its location on a map. The Moon has no Global Positioning System (GPS) to provide constant, precise location data. Instead, the lander must figure out its position, altitude, and velocity entirely on its own. This is accomplished through a complex process called sensor fusion.

The lander’s flight computer constantly ingests data from multiple sources: Inertial Measurement Units (IMUs) track acceleration and rotation, star trackers determine the spacecraft’s orientation against the backdrop of deep space, and onboard cameras perform optical navigation by taking pictures of the surface and matching crater patterns to pre-loaded maps. During the final descent, radar or laser altimeters (lidar) provide critical, high-precision measurements of the distance to the ground and the speed of descent. A failure in any one of these sensors, or in the software that weaves their data into a coherent picture, can lead the lander astray. The system must be robust enough to handle conflicting information or noisy sensor readings, a lesson learned the hard way by several recent missions that suffered from altimeter-related issues.

The Tyranny of Gravity and Vacuum

On Earth, a descending vehicle can use the friction of the atmosphere and parachutes to slow down. The Moon, however, is in a near-perfect vacuum, which means there is no air to brake against. The entire deceleration—from an orbital speed of over 1,600 meters per second to a gentle, controlled touchdown—must be performed using only rocket engines.

This requires a main engine that can be precisely throttled up and down, along with a set of smaller attitude control thrusters to keep the lander stable and correctly oriented. The entire landing is a continuous, powered fight against the Moon’s gravity. Fuel margins are incredibly tight to save weight on the long journey from Earth. Any unexpected event, such as an engine that produces slightly more or less thrust than commanded, a navigation error that requires a course correction, or the need to hover for extra seconds to find a safe landing spot, can rapidly deplete the precious fuel reserves. This was a key factor in the failure of Japan’s Hakuto-R Mission 1, which hovered at an incorrect altitude until its tanks ran dry, sending it plummeting to the surface.

An Uncharted and Treacherous Landscape

The lunar surface is not a benign, flat plain. It is a rugged, ancient world covered in craters of all sizes, fields of boulders, and steep, unstable slopes. The polar regions, which are of intense scientific interest due to the potential for water ice in permanently shadowed craters, are particularly treacherous.

Modern landers must be equipped with sophisticated hazard detection and avoidance systems. In the final seconds before touchdown, the lander’s computer must analyze the ground below, identify any dangers, and autonomously command the thrusters to divert to a safer spot if the original target is unsuitable. This process is made even more difficult by the extreme lighting conditions at the poles. The sun hangs low on the horizon, casting long, sharp shadows that can hide craters and boulders or confuse optical navigation systems. An area that appears safe in high-resolution orbital maps can look very different to the sensors of a descending spacecraft. This lighting was a contributing factor in the partial failure of the IM-2 mission, which tipped over after landing. Furthermore, the lander’s own rocket exhaust can create a blinding cloud of dust and gravel, which can obscure sensors and damage the spacecraft itself.

The Final Minutes: A High-Stakes Autonomous Operation

Perhaps the most daunting challenge is that the entire landing must happen without any direct human control. It takes a radio signal about 1.3 seconds to travel from Earth to the Moon, meaning the round-trip communication time is nearly three seconds. This delay makes it impossible for a mission controller on Earth to pilot the lander in real time.

The spacecraft is completely on its own during the “20 minutes of terror” of the landing sequence. Every action is governed by pre-programmed software running on the onboard flight computer. This single computer must flawlessly manage the complex interplay of navigation sensors, propulsion systems, and hazard avoidance algorithms. A software bug, a processor that lags, or an unexpected sensor reading that the software isn’t programmed to handle can lead the computer to make a fatal error. This was the root cause of Russia’s Luna 25 crash, where a control unit failed to issue the command to shut down the engine, and the first Hakuto-R crash, where the software misinterpreted altimeter data and commanded the lander to perform an impossible maneuver.

A New Lunar Era: Missions and Outcomes (2023-2025)

The period between 2023 and 2025 served as a crucible for this new generation of lunar explorers. The outcomes of these missions, summarized below, provide a clear narrative of the challenges, the learning process, and the gradual progress being made.

2023 – A Year of Contrasting Fortunes

The year 2023 encapsulated the high-risk, high-reward nature of lunar exploration. It began with the ambitious attempt of a private Japanese company, ispace, with its Hakuto-R Mission 1. The mission aimed to be the first privately funded spacecraft to land on the Moon. It successfully reached lunar orbit, but the landing attempt on April 25 ended in failure. Analysis revealed a software flaw was the culprit. As the lander passed over the rim of the Atlas crater, its altimeter correctly registered the sudden change in altitude. The flight software, however, incorrectly interpreted this valid data as a sensor error, ignored it, and concluded the lander had already touched down when it was still 5 km above the surface. The lander then hovered at this incorrect altitude, burning its remaining propellant until it ran out and plummeted to the surface.

In August, the world watched as Russia attempted its first lunar mission in 47 years. The Luna 25 lander was on a fast track to the Moon’s south pole, aiming to land just days before India’s competing mission. On August 19, during a maneuver to adjust its orbit in preparation for landing, the probe’s engine fired for 127 seconds instead of the planned 84. The extended burn pushed the spacecraft onto a trajectory that intersected with the Moon. Roscosmos later reported that the error was caused by an onboard control unit that failed to shut down the engine because it did not receive the necessary data from an accelerometer, which had not been switched on due to a command priority issue. The probe crashed into the Pontécoulant G crater, ending Russia’s return to the Moon in failure.

Just four days after the Luna 25 crash, on August 23, the Indian Space Research Organisation (ISRO) achieved a historic triumph. Its Chandrayaan-3 mission successfully executed a flawless soft landing near the lunar south pole, making India only the fourth nation to land on the Moon and the very first to reach its coveted southern polar region. The mission, consisting of the Vikram lander and the small Pragyan rover, demonstrated a mastery of the landing sequence that had eluded its predecessor, Chandrayaan-2, in 2019. The success was celebrated globally as a major milestone in space exploration.

2024 – The Commercial Proving Ground

The year 2024 was defined by the first missions under NASA‘s Commercial Lunar Payload Services (CLPS) program, which relies on private companies to deliver science to the Moon. The first to launch, on January 8, was Astrobotic’s Peregrine Mission 1. Hopes were high, but just hours after a successful launch, the mission was in jeopardy. A valve in the propulsion system failed to reseal after activation, causing high-pressure helium to rupture an oxidizer tank. The resulting critical propellant leak made a lunar landing impossible. After operating in space for over a week and collecting valuable data from its powered-on payloads, the Astrobotic team guided the crippled spacecraft to a controlled and safe reentry into Earth’s atmosphere to avoid creating space debris.

Days after Peregrine’s mission ended, Japan’s space agency, JAXA, achieved a remarkable success with its SLIM (Smart Lander for Investigating Moon) mission on January 19. The primary goal was not just to land, but to demonstrate “pinpoint” landing technology. SLIM achieved this with stunning accuracy, touching down just 55 meters from its target point on the slope of Shioli crater. The landing was not without issue; a problem with one of its main engines caused the lightweight lander to tip over onto its nose, pointing its solar panels away from the sun. It operated on battery power for a few hours before being shut down. In a stunning turn of events, as the sun’s angle changed over the following week, light reached the solar cells, and JAXA was able to “resurrect” the lander. Even more remarkably, SLIM, which was not designed to survive the frigid lunar night, subsequently woke up after multiple two-week periods of deep cold, continuing to send back data and images.

The first successful commercial landing finally came on February 22, when Intuitive Machines’ IM-1 lander, named Odysseus, touched down near the Malapert A crater. It was the first American spacecraft to soft-land on the Moon since Apollo 17 in 1972. The landing was a dramatic, high-stakes improvisation. During the final approach, the team discovered that the lander’s primary laser rangefinders were not working. In a remarkable feat of real-time problem-solving, engineers on the ground wrote and uploaded a software patch to enable the lander to use an experimental NASA lidar payload as its primary navigation sensor. The workaround was successful, and Odysseus landed softly. However, it came in with some horizontal velocity, breaking a landing leg and tipping over onto its side, which limited the effectiveness of some payloads and its communication antennas.

2025 – Learning Curves and Breakthroughs

The lessons from the previous years began to bear fruit in 2025. On March 2, Firefly Aerospace’s Blue Ghost Mission 1 achieved what no commercial company had before: a fully successful, upright landing on the Moon. The lander touched down in Mare Crisium and proceeded to carry out its mission flawlessly. It successfully operated all 10 of its NASA-funded science payloads, ran for its entire planned mission of one lunar day (14 Earth days), and continued to operate for over five hours into the darkness of the lunar night, setting a new standard for commercial mission success.

Just days later, on March 6, Intuitive Machines made its second attempt with the IM-2 lander, named Athena, targeting the extremely challenging terrain of Mons Mouton near the south pole. The mission ended with a similar outcome to its predecessor: a partial success. The lander reached the surface intact but, like Odysseus, tipped over onto its side. An investigation revealed a combination of factors: signal noise and distortion interfered with the laser altimeter’s readings, and the difficult low-angle sunlight and long shadows at the south pole confused the optical navigation system, which relies on recognizing crater patterns from orbital maps.

The year also saw a second failure for the Japanese company ispace. Its Hakuto-R Mission 2, named Resilience, attempted a landing on June 5. The mission ended in another hard landing. The company’s analysis determined the cause was different from its first attempt. This time, the failure was attributed to a hardware fault in a new Laser Range Finder (LRF). The sensor was slow to provide accurate altitude data during the descent, causing the flight computer to underestimate its speed and fail to decelerate sufficiently.

Case Studies in Success and Failure

Moving beyond a simple timeline, a comparative analysis of these missions reveals critical patterns in engineering philosophy, technological dependencies, and the definition of success itself.

Success Through Iteration: The Chandrayaan-3 Story

The triumph of India’s Chandrayaan-3 was not an isolated event but the culmination of a deliberate and transparent learning process following the 2019 failure of Chandrayaan-2. ISRO adopted what its chief called a “failure-based design” philosophy, meticulously analyzing what went wrong with the previous mission and systematically engineering solutions to prevent those failures from recurring.

Several key upgrades were implemented. The lander’s legs were strengthened to withstand a higher landing velocity, increasing from 2 m/s to 3 m/s. More fuel was added to provide a larger margin for course corrections and hovering. To improve navigation, a Laser Doppler Velocimeter (LDV) was added, providing a redundant and dissimilar way to measure the lander’s velocity in three directions. The target landing zone was expanded from a tight 500m-by-500m square to a more forgiving 4km-by-2.5km rectangle, giving the autonomous system more flexibility to find a safe spot. Finally, additional solar panels were placed on other surfaces of the lander to ensure it could generate power even if it did not land perfectly upright. This methodical approach, which directly addressed previous shortcomings, proved highly effective and stands as a powerful example of successful iteration in spacecraft design.

The Altimeter’s Curse: A Tale of Two Companies

The missions of ispace and Intuitive Machines highlight a common vulnerability in modern lunar landers: the critical dependence on accurate altitude and velocity data. The failures and partial failures of these companies can be traced back to what can be called the “altimeter’s curse”—a breakdown in the chain of information from the sensors that measure distance to the ground to the flight computer that acts on that data.

For ispace, the two failures had different root causes but the same result. In Mission 1, the hardware worked, but the software misinterpreted the data. In Mission 2, the software was presumably correct, but the new Laser Range Finder (LRF) hardware itself failed to perform as expected, delivering data too late for the lander to react. In both cases, the lander was effectively flying blind, unaware of its true height above the treacherous lunar surface.

Intuitive Machines faced a similar challenge. On IM-1, the primary laser altimeters were rendered useless by a pre-launch human error (a safety switch was not disabled), forcing the risky, last-minute switch to a NASA payload. On IM-2, the altimeters were active but were confused by signal noise and the unique lighting conditions at the south pole. While both landers managed to touch down softly, the inaccurate velocity and altitude data contributed to their sideways landings. This pattern reveals that the sensor-to-software interface is a modern Achilles’ heel. It’s not enough to simply have the sensors; the entire autonomous system must be robust enough to handle the imperfect, noisy, or even absent data that is a reality of operating in a harsh and unpredictable environment. These incidents suggest a potential gap in the ability of ground-based testing to fully replicate the complex interplay of hardware, software, and the lunar environment.

Precision and Resilience: JAXA’s SLIM and Firefly’s Blue Ghost

The most impressive successes of the period, from JAXA and Firefly Aerospace, demonstrate two different but equally vital pathways to success in the new lunar era.

JAXA’s SLIM mission was a triumph of technological demonstration. Its primary goal was to prove that a lander could achieve “pinpoint” accuracy. It did so spectacularly, using a novel vision-based navigation system that continuously compared live images with stored crater maps to guide itself to a landing within meters of its target. This is a revolutionary capability that will allow future missions to target specific, scientifically interesting features—like the edge of a crater or a unique geological formation—with unprecedented precision. SLIM’s success was compounded by its unexpected resilience. Despite landing on its nose and not being designed for it, the probe survived multiple two-week-long, -170°C lunar nights, showcasing a remarkable robustness.

In contrast, Firefly’s Blue Ghost mission was a triumph of commercial service delivery. While it did not attempt the same level of pinpoint accuracy as SLIM, its success was defined by its end-to-end operational perfection. The lander performed an upright, stable landing, successfully deployed and operated all of its customer payloads, met 100% of its mission objectives, and exceeded its planned operational lifetime. It functioned exactly as a reliable delivery service should.

These two missions highlight a crucial distinction. SLIM’s success was in advancing the technological frontier, demonstrating what is possible. Blue Ghost’s success was in demonstrating that lunar landing can be a reliable, repeatable commercial service. Both are essential for building a sustainable presence on the Moon. One pushes the boundaries of exploration, while the other builds the dependable logistics needed to support it.

The Path Forward: Lessons for the Artemis Generation

The intense activity of 2023-2025 is not happening in a vacuum. It is directly feeding into NASA‘s Artemis program and the broader goal of establishing a long-term human presence on the Moon. The lessons learned from these robotic missions are shaping the technologies and strategies that will be used for decades to come.

The CLPS Model: High Risk, High Reward

The missions from Astrobotic, Intuitive Machines, and Firefly are all part of NASA‘s Commercial Lunar Payload Services (CLPS) initiative. This program represents a fundamental shift in how NASA conducts business. Instead of managing every aspect of a mission, NASA acts as a customer, buying a delivery service from a commercial provider.

The core philosophy of CLPS is to accept a higher level of risk in exchange for lower costs and a much faster mission cadence. NASA understands that with this model, some commercial missions will fail. The period from 2024-2025 shows this strategy playing out in real time. The failure of Peregrine and the partial successes of the IM landers are part of the accepted cost of doing business. The data gathered from these missions, even the failed ones, is invaluable. In return, the full success of Firefly’s Blue Ghost demonstrates the enormous potential of the program: a successful science delivery mission at a fraction of the cost and time of a traditional, government-led project. This rapid succession of flights creates a fast-paced learning environment that is accelerating innovation across the industry.

The Future of Lunar Exploration

The lessons from this period are already shaping the future. Despite the high-profile failure of Luna 25, Russia has affirmed its commitment to its lunar program, announcing plans for future Luna missions and a potential collaboration with China on projects like a lunar nuclear power station.

The commercial sector has proven its resilience. Companies like ispace and Intuitive Machines are not retreating after their setbacks. They are actively incorporating the hard-won lessons into their next-generation landers. The technological evolution is clear: future landers will feature more robust and redundant systems. Intuitive Machines, for instance, is adding dissimilar altimeters, a lighting-independent sensor for measuring velocity, and an improved crater database to its software for its next flight.

This period marks a fundamental shift from the “proof of concept” era of Apollo to the “operational logistics” era of Artemis. The question is no longer simply “can we land on the Moon?” but rather “can we land reliably, precisely, and cost-effectively, wherever we need to, every single time?” The mixed results of the last few years show that the industry is still in the early, volatile stages of answering that question. The successes are demonstrating the path forward, while the failures are mapping the pitfalls to avoid. This period will be remembered as the crucible in which the foundational capabilities for a sustainable human return to the Moon were forged.

Summary

Returning to the Moon and landing safely on its surface remains an exceptionally difficult task, defined by a series of unforgiving challenges. A lander must navigate autonomously without GPS, use its engines to fight gravity in a vacuum, and find a safe haven in a treacherous and poorly lit landscape, all without real-time human guidance.

The period between 2023 and 2025 put these challenges on full display. A series of missions from national agencies and new commercial companies met with a mixture of stunning success and heartbreaking failure. Early attempts by Russia’s Roscosmos and Japan’s ispace ended in crashes caused by engine control and software errors. In contrast, India’s ISRO achieved a historic success with Chandrayaan-3, learning directly from a previous failure to become the first nation to land near the lunar south pole.

In the commercial sector, NASA‘s CLPS program saw its first missions take flight, yielding a spectrum of results. Astrobotic’s Peregrine lander was lost to a propellant leak, while Intuitive Machines’ IM-1 and IM-2 landers reached the surface but tipped over due to persistent issues with their altimetry systems. Japan’s JAXA demonstrated revolutionary pinpoint landing technology with its SLIM lander, and Firefly Aerospace achieved the first fully successful end-to-end commercial mission with its Blue Ghost lander.

The core lesson from this intense period of activity is that success in the modern lunar era hinges on creating robust, resilient, and intelligent autonomous systems. The most effective path to developing these systems is through an iterative process of building, flying, and learning from both the triumphs and the failures. This is the difficult but necessary work that is paving the way for a sustained human presence on the Moon and beyond.