Three rockets, three upgrade paths to the Moon
Getting American astronauts back to the lunar surface now depends on three rockets that mostly do not look alike, and on the fact that all three are changing at once. On 23 October 2026, at a press conference at Kennedy Space Center, NASA laid out how Artemis III will be flown, and the answer was a rehearsal, not a landing. Four astronauts will launch on the Space Launch System in 2027, dock with a test Blue Moon lander and a test Starship in Earth orbit, and come home. The landing itself moves to Artemis IV in 2028, and the first crew on the surface since Apollo 17 will ride on whichever of the two commercial landers proves ready first (NASA's release on the revised plan).
That rework set off a hardware sprint across three launch pads. Each vehicle is getting a new upper stage, a new engine, or a whole new configuration, and the changes are not interchangeable. SLS is losing the upgrade it was supposed to get and gaining a different one. Starship is getting a docking adapter, a tanker fleet and an upper stage for its own lander. New Glenn is getting a bigger first stage, a rebuilt pad and a four-launch plan that was not in the original design. Side by side, the three are three answers to one question: how do you get a crew to the Moon and back without flying the same hardware twice?
The reordering also exposes an uncomfortable schedule. NASA Administrator Jared Isaacman announced the new sequence roughly a month before Artemis II lifted off in April 2026, and officials now describe Artemis III as an Earth-orbit test in mid-2027, with Artemis IV and Artemis V both aimed at landings in 2028 (SpacePolicyOnline on the schedule). The South Pole remains the destination, justified by water ice thought to be accessible in permanently shadowed terrain, though the agency concedes it needs more data from robotic landers before committing. The suits are being built by Axiom Space. One detail is significant: the docking test happens in low Earth orbit rather than near the Moon, letting NASA learn the hard parts of rendezvous while a crew is still a quick flight from the splashdown.
The Space Launch System is the odd one out, the only one of the three being simplified. NASA's original plan was a Block 1 rocket with the Interim Cryogenic Propulsion Stage for the first missions, then a Block 1B with Boeing's larger Exploration Upper Stage and a Block 2 with new solid boosters. That ladder was dropped in February 2026; the agency settled on a standardized configuration and picked ULA's Centaur V as the upper stage to replace the interim stage from Artemis IV onward (Space.com on the upper-stage selection). Centaur V has flown on Vulcan since 2024 and holds roughly twice the propellant of the interim stage, and NASA is working with ULA on an extended version for the low-lunar-orbit role the new architecture calls for (Ars Technica on the revised architectures). For Artemis III itself, SLS will fly without a propulsion stage at all, using a non-propulsive spacer instead, since the mission never leaves Earth orbit.
Starship's contribution to the same plan is the most expensive one, and the most contested. Under the revised concept, tanker variants of Starship fill a depot ship in low Earth orbit, the lander and Orion dock there, and the lander performs the burn that sends the combined stack toward the Moon (NASA's release on the revised plan). How many tankers that takes is the argument. At a NASA Aerospace Safety Advisory Panel meeting, panel member and former astronaut Charlie Precourt put the estimate at around 22 flights, noting the large number of firsts stacked against Artemis IV. Elon Musk responded that the real figure would be closer to six to eight (Ars Technica on the tanker estimate). The gap is not just numerical. It depends on how much cryogenic methane and liquid oxygen boil off in orbit and how efficiently the transfer works, neither yet demonstrated. Even at a twice-weekly launch rate by the end of 2027, Ars Technica's analysis puts the time to fuel a single depot at roughly three months.
New Glenn is the third rocket, and it is carrying the most physical damage. On 28 May 2026, a static fire at Cape Canaveral destroyed a fully stacked New Glenn and much of Launch Complex 36. Blue Origin traced the anomaly to the main liquid oxygen valve on one of the seven BE-4 engines in the first stage, with the root cause still open at the time of reporting, and began rebuilding the pad in June with a public target of returning to flight before the end of 2026 (KeepTrack on the LC-36 explosion and recovery). The rebuilt pad, LC-36A, flies the current 7x2 configuration, while LC-36B is being built for the heavier 9x4. New Glenn has flown three times, its third on 19 April 2026 ending with the loss of the payload, and its November 2025 ESCAPADE booster reused on that attempt. All of that sits underneath a lander campaign that needs the rocket to fly four times.
The engineering behind the new variants is where the three programs diverge most sharply. New Glenn 9x4 replaces the current 7x2 with nine BE-4 engines on the first stage and four BE-3Us on the upper stage, carries a fairing widened from 7 metres to 8.7 metres, and lifts roughly 70 metric tons to low Earth orbit against 45 tons for the original configuration, with the engine upgrades phased in starting with the third mission (Space.com on the 9x4 variant). Starship's lunar lander and its depot ship shed hardware rather than adding it: neither re-enters Earth's atmosphere, so both drop the heat shield and the fins, and the lander keeps six Raptor engines for ascent and landing while separate high-thrust engines in the mid-body burn gaseous oxygen and methane to avoid blasting the regolith (Wikipedia on the Starship HLS vehicle family). Blue Moon takes the opposite approach, keeping liquid hydrogen and liquid oxygen with a solar-powered system that holds propellant near 20 kelvin to fight boiloff, which matters for a lander meant to sit on the Moon rather than leave (Wikipedia on the Blue Moon lander family). And beneath all of it, the crew rocket changes hands: from Artemis V, NASA plans to transfer SLS production and launch operations to Deep Space Transport LLC, a Boeing and Northrop Grumman joint venture, hoping commercial customers spread the cost (Wikipedia on Deep Space Transport LLC).
The three vehicles have also started flying. Starship reached orbit for the first time on 28 September 2026 on flight test 14, deploying 26 operational Starlink V3 satellites after one of the six second-stage engines shut down early, which cut the planned six-orbit profile short but not the orbit itself (SpacePolicyOnline on the first orbital flight). For Artemis III, the sequence runs to a 14-day nominal mission: New Glenn goes first with the Blue Moon test article, SLS launches the crew next, and Starship flies on about flight day six to dock on day eight, with the Blue Moon docking on day three at roughly 273 miles (Space.com on the concept of operations).
What ties the three together is the 2028 landing, and what separates them is how each program pays for it. Blue Origin chose the conservative route, deferring in-orbit propellant transfer and instead flying three transfer stages built from existing Mark 1 and Mark 2 hardware ahead of the Blue Moon Mark 2 Alpha lander, which assembles in Earth orbit and departs for an elliptical polar orbit around the Moon. That is four New Glenn launches with no spacecraft-to-spacecraft refueling, on a booster whose pad was vaporized in May. SpaceX chose the opposite, betting that orbital refueling works well enough that a handful of tanker flights will cover the lander, a bet its own safety advisory board thinks is off by more than a factor of three. NASA bought optionality by choosing neither, flying both dockings in Earth orbit in 2027 and deciding the landing provider from what the data shows. The next real milestone is not a landing but a docking, and the hardware that has to make it work is still in pads and test stands.
NASA's Artemis III operations concept graphic, showing the three launch vehicles and the docking sequence in low Earth orbit.
Featured blog posts
A Transformer model called EarlyDetect, trained on SDO/HMI helioseismic and magnetic observations by an NJIT-led team including Princeton and NASA Ames, forecasts the emergence of solar active regions on the Sun an average of 9.24 hours before they become visible, outperforming LSTM baselines and revealing that temporal smoothing filters hid the faintest precursor signals (Tirona et al., JGR Machine Learning and Computation, 14 August 2026) 
Sixty-three relays over your head: how the U.S. Space Development Agency quietly built a sky internet. Posted on May 24, 2026
New observations from the James Webb Space Telescope are providing the clearest and most detailed view yet of the environment where these carbon structures form and evolve. 
Engineers and scientists are assembling a spacecraft designed for a specific and increasingly important purpose: finding potentially hazardous objects before they find Earth. Posted on May 4, 2026
An international team of researchers announced that NASA's Curiosity rover had identified more than 20 distinct organic molecules preserved in ancient Martian rocks. 
To truly shorten the distances between planets, something more powerful is required—something that does not merely burn fuel, but transforms matter itself into energy. 
There are moments in engineering when progress is obvious. And then there are moments when progress looks like subtraction. 
For as long as humans have pushed aircraft beyond the speed of sound, there has been a cost to that achievement—an invisible but unmistakable shockwave that ripples across the sky and crashes into the ground as a sonic boom. 
There are moments in the history of technology when an idea appears so simple in form and so vast in implication that it changes how we think about the future. 
The American Institute of Aeronautics and Astronautics has released a groundbreaking report identifying ten technologies that will fundamentally reshape aerospace. 
Every propulsion revolution has been delayed not by imagination, but by power. 
The LASER taught us that energy does not need to explode outward to be useful. It can be channeled. It can be persuaded to exit matter in an orderly way. 
The story of the LASER does not begin with light, but with order. 
Today we are joined byYasunori Yamazaki, Chief Business Officer at Axelspace. Axelspace are pioneers of microsatellite technology advancing the frontiers of space business, reimagining traditional ways of using space, and creating a society where everyone on our planet can make space part of their life. 
Today we are joined by Logan Ryan Golema, Founder & Principal, and Vishal Singh, Chief Scientist at Lunargistics. Latest blog posts
NASA's October 23, 2026 Artemis reset puts SLS, SpaceX's Starship and Blue Origin's New Glenn into a 2027 Earth-orbit docking test, and each of the three launch vehicles is getting a new upper stage, a new engine or a new variant to make the 2028 landing work 
NASA put four volunteers in a 1,700-square-foot 3D-printed habitat at Johnson Space Center for 378 days, with 22-minute one-way radio delays and equipment that fails on schedule, to measure what isolation and a year of waiting do to a crew. 
The Defense Intelligence Agency's 2026 Challenges to Security in Space report publishes the first full-resolution public image of China's reusable Shenlong spaceplane, captured in orbit in May 2024 by a Maxar (now Vantor) WorldView-3 commercial satellite, showing delta wings, a payload bay, a dark heat shield and a rear service module with solar arrays, alongside a flight record of four missions since 2020 and the agency's finding that China may field space-based directed-energy weapons by the late 2020s 
MIT engineers grew muscle cells less than 15 micrometres thick onto a half-millimetre gel film and got a pair of fins to pull a robot through water at four body lengths per minute. Freed from the stiff backing that used to kill thin muscle within minutes, the same film kept actuating for more than a month. 
What proof-of-work consensus faces once a cryptocurrency network stretches across the solar system: light delay breaks the chain long before relativistic time dilation does, and the workable answers are local settlement layers, cross-world bridges, and a single coordinate clock for every node 
The spacecraft's two main-belt flybys have already paid for themselves: one asteroid turned out to have a satellite that is itself a contact binary, the other a peanut-shaped body tumbling on more than one axis. The first close encounter with a real target is set for 11 August 2027. 
The Space Force pulled six military missions off ULA's Vulcan after one of its solid rocket motor nozzles melted through in February. The November 13 return flight carries 40 Amazon Leo satellites and, for the first time, six strap-on boosters. 
Northwestern physicists held one antiparticle in a cryogenic Penning trap, cooled it to its cyclotron ground state, and resolved individual quantum jumps for months on end. The setup makes an electron-antiparticle magnetic-moment comparison precise enough to put CPT symmetry under a level of scrutiny the lepton sector has never seen. 
Adding the four loss channels a real reactor cannot avoid shifts the least demanding path to ignition away from the one most designs plan, and a trace of tungsten eroded from the walls can double the pressure a machine would need. 
summary: "OpenAI and Anthropic each claim a model produced a deterministic algorithm that decides 3-SAT in subexponential time, which would knock out the Exponential Time Hypothesis and the conditional hardness results built on it. Both ship Lean formalizations, neither has been peer reviewed, and the bounds only start to beat brute force on formulas with more than 10^100,000,000 variables. 
NASA-funded researchers report that Teghaza 001, a coarse-grained Martian gabbro found in the Mali Sahara in 2022, formed at least 4.11 billion years ago, making it a sample of Martian early crust and mantle from an era no other returned rock covers, with uridium-lead dating of its zircon crystals and in-situ rubidium-strontium analysis giving consistent ages near 4.1 and 4.24 billion years, oxygen-18 enrichment pointing to water-rock interaction, and a deuterium-to-hydrogen floor of roughly twice Earth's seawater ratio implying Mars lost hydrogen to space early 
The Air Force Research Laboratory's first public Rocket Cargo experiment put cargo pods on a commercial SpaceX rideshare, then recovered them from the Pacific with sensors logging the temperature, shock and vibration that fragile equipment has to survive. 
After eight years and 10 billion kilometres of spiralling inward, the ESA-JAXA mission fires its most consequential burn on 21 November 2026: a single automated capture into orbit around Mercury that controllers cannot correct in real time. 
A newly assigned closed-cycle converter moves the Mars mission's most distinctive subsystem from concept to contracted hardware. The plan puts a 20-kilowatt-electric fission system on a 2028 launch, with three Mars helicopters as payload and no reactor operation during launch. 
NASA's October 23, 2026 plan turns the first Artemis landing into a rehearsal: four astronauts will orbit Earth in 2027 and dock with a Blue Origin lander test article and a SpaceX Starship before any crew returns to the Moon, now targeted for early 2028. 
Subscribe to blog posts using RSS
