28 May 2026
The global aerospace community released a sweeping new assessment on the state of spaceflight, concluding that while the technical challenge of launching rockets has been largely mastered, the landing phase remains the single greatest threat to mission success. According to the International Consortium for Atmospheric and Reentry Research (ICARR), which published its annual “Return to Earth” report today, over 40 percent of all launch vehicle and spacecraft failures since 2020 have occurred during descent and terminal landing, a statistic that has remained stubbornly unchanged despite dramatic advances in propulsion and guidance systems. “We have essentially solved the problem of getting off the ground reliably,” said Dr. Helena Voss, ICARR’s lead propulsion analyst. “But bringing a vehicle back intact, whether it is a reusable booster or a crew capsule, still involves a cascade of chaotic variables—atmospheric buffeting, thermal gradients, and split-second actuation of landing gear or parachutes—that our models cannot fully tame.”
The report highlights a stark divergence in progress: orbital launch success rates now exceed 97 percent across major spacefaring nations, driven by mature liquid-engine designs, automated health checks, and decades of operational refinement. However, landing success for reusable first stages hovers near 82 percent, and for capsules returning from the Moon or Mars, the figure drops below 70 percent when accounting for partial anomalies. “People assume that because we see boosters landing on droneships on social media, the hard part is over,” remarked Professor Kenji Tanaka of the Tokyo Institute of Space Dynamics, who contributed data on hypersonic reentry. “In reality, each landing is a controlled crash that must decelerate from Mach 25 to zero in under ten minutes. One stuck grid fin, one unexpected wind shear, one sensor lag, and the vehicle becomes a fireball. The margin for error is millimeters and milliseconds.”
The 28 May briefing also reviewed two recent near-disasters that underscore the fragility of current landing systems. On 14 April 2026, a commercial Terran R rocket successfully delivered a satellite to geostationary transfer orbit, but during its return to the launch site, a hydraulic pump failure caused three of its four landing legs to deploy asymmetrically. The booster touched down at a 12‑degree tilt, skidding off the pad before autonomous safing systems shut down its engines. Although no personnel were injured, the vehicle was declared a total loss. Weeks earlier, on 3 May, a Russian Orel crew capsule returning from the International Space Station experienced a partial failure of its soft-landing thrusters just 50 meters above the Kazakh steppe.
The capsule struck the ground at 8 meters per second—twice the design limit—resulting in minor spinal compression for one cosmonaut and a six‑month grounding of the Orel fleet. “These are not freak accidents,” said Dr. Voss. “They are statistical certainties given how many complex systems must work in perfect harmony at the very moment when the vehicle is most vulnerable. Launch has redundancy and abort modes. Landing often has none.”
Beyond engineering hurdles, the report emphasizes a looming regulatory and insurance crisis. Premiums for landing‑based mission phases have risen 300 percent since 2023, and several underwriters now refuse to cover reusable first‑stage recovery unless the operator flies at least five consecutive successful landings of the same vehicle type. “The economics of reusability depend entirely on landing reliability,” explained Maria Flores, CEO of the launch logistics firm Ascent Analytics. “If you lose one in five boosters, you might as well throw them away after each flight. We are at a tipping point where the industry’s obsession with landing must shift from ‘proving it can be done’ to ‘making it as boring as a commercial airliner touchdown.’ That requires a complete rethinking of landing control laws, sensor fusion, and perhaps even mechanical simplicity over software complexity.”
The report’s authors call for a dedicated international “Landing Safety Initiative” to mirror the 1970s‑era work on launch abort systems. Recommended priorities include real‑time wind‑shear detection using vehicle‑mounted LIDAR, redundant, mechanically independent landing leg actuators, and AI‑driven flight computers that can land a vehicle even with multiple thruster or aerodynamic surface failures. Several space agencies have already signaled interest.
The European Space Agency announced today that its Themis reusable rocket demonstrator will be refitted with a triple‑redundant leg system and tested in high‑sea states off French Guiana starting September 2026. Meanwhile, NASA’s Human Landing System program for the Artemis Moon missions has quietly added a new requirement: any lunar lander must demonstrate two consecutive successful landing aborts from 100 meters altitude before being certified for crewed flight. “The Moon is even harder than Earth,” noted Professor Tanaka. “No atmosphere, no parachutes, just a rocket engine and a hope that the radar altimeter doesn’t mistake a boulder for a flat surface. If we don’t fix landing on Earth first, we have no business trying it on another world.”
As commercial space stations, lunar cargo missions, and Mars sample return plans accelerate toward the late 2020s, the 28 May report serves as a sobering reminder: launching is an act of power; landing is an act of precision. And precision, unlike brute force, cannot be brute‑forced. “We have become complacent because rockets go up beautifully every week,” concluded Dr. Voss. “But until we treat the landing with the same obsessive rigor as the ascent, every mission ends with a bet against physics. And physics always wins in the end.” The consortium’s full 450‑page technical annex, released simultaneously, will be discussed at the International Astronautical Congress in October 2026—provided, as one engineer joked, that participants can land their aircraft safely.
