The US National Aeronautics and Space Administration (NASA) has abandoned plans to create a Lunar Gateway orbital station around the Moon to focus on building a base at the lunar South Pole. To this end, NASA is accelerating its Artemis programme.
Announcing the decision, NASA Administrator Jared Isaacman said the agency “is suspending Gateway in its current form and will focus on the infrastructure necessary for the long-term presence of people on the Moon. The cost of the new lunar base is estimated at approximately $20bn.
Gateway was originally conceived as an orbital station from which astronauts would transfer to landing modules before descending to the Moon. The project was already at an advanced stage of development involving contractors Northrop Grumman and Lanteris Space Systems.
Now the station’s equipment and developments are planned to be used to create ground infrastructure. This complicates the roles of international partners – Japan, Canada and Europe – which initially developed modules for the orbital station.
The changes affect the entire architecture of the Artemis programme, including multi-billion-dollar contracts with private contractors. In particular, SpaceX (founded by Elon Musk) and Blue Origin (founded by Jeff Bezos) continue to develop manned lunar modules, although both are behind schedule.
The Artemis programme, launched in 2017, aims for manned moon landings as the basis for regular missions for the first time since the Apollo programme, which ended in 1972. Artemis I was the inaugural uncrewed flight test between November and December 2022 that served as a critical demonstration to certify the Space Launch System (SLS) rocket and the Orion spacecraft for future crewed missions.
The most immediate changes of the revised programme redefine Artemis III, which will no longer land on the Moon. Artemis II (targeted for April 2026) remains a crewed 10-day lunar flyby. The SLS rocket is currently at the Kennedy Space Center preparing for launch.
Artemis III (mid-2027) has been re-tasked as a Low Earth Orbit (LEO) demonstration mission. The Crew will practice docking the Orion capsule with one or both commercial landers (SpaceX’s Starship or Blue Origin’s Blue Moon) to reduce technical risk. Artemis IV (early 2028) is now designated as the first crewed lunar landing mission while Artemis V (late 2028) is planned as a second landing later the same year to establish an annual surface mission cadence.
NASA is increasing competitive pressure on its contractors due to schedule slips. NASA will no longer stick to a fixed provider sequence. Either SpaceX or Blue Origin could perform the first landing in 2028, depending on which vehicle passes the 2027 LEO docking tests. Beyond Artemis V, NASA intends to phase out the government-owned SLS in favour of commercially procured, reusable rockets to enable landings as frequently as every six months.
NASA has shifted its architectural goals to compete more aggressively with the International Lunar Research Station (ILRS) – a comprehensive scientific base at the Moon’s South Pole, co-led by China and Russia aiming for a basic crewed version by 2035. “The clock is running in this great‑power competition, and success or failure will be measured in months, not years,” said Isaacman.
The ILRS architecture differs from NASA’s by prioritising autonomous, heavy-duty “hard power” systems. China and Russia are collaborating on a lunar nuclear reactor intended to provide a constant energy source for labs and pressurised rovers, ensuring the base remains operational during the 14-day lunar night.
NASA previously confirmed plans to deploy a nuclear fission reactor on the lunar surface by 2030 through the Fission Surface Power (FSP) project. The target is a 100 kW reactor that can operate for at least 10 years. It is designed to be relatively small – fitting within a 6,000 kg launch mass to be transported by commercial heavy-lift rockets. NASA and the Department of Energy (DOE) are working with commercial partners to finalise designs by early 2026.
In addition to outlining the reconfiguration of the Artemis programme, NASA announced that it is preparing the Space Reactor-1 (SR-1) Freedom mission, a device with a nuclear electric propulsion system, which is planned to be sent to Mars by the end of 2028. “With SR‑1 Freedom, we are finally putting nuclear propulsion on a trajectory out of the laboratory and into deep space,” said NASA Associate Administrator Amit Kshatriya.
The spacecraft is a hybrid of repurposed legacy hardware and cutting-edge nuclear technology. Unlike nuclear thermal rockets that use heat to expand propellant, SR-1 Freedom will use a 20-kW fission reactor to generate electricity. This powers high-efficiency ion thrusters, allowing for sustained thrust and massive payload transport. The reactor will use high-assay low-enriched uranium (HALEU) oxide fuel, heat pipes for thermal transfer and a boron carbide shield to protect onboard electronics from radiation. Excess heat is managed via a massive composite and titanium heat sink. The reactor is placed at one end of a long truss, with sensitive electronics and the propulsion system at the far end to minimise radiation interference.
The data gathered from SR-1 Freedom will directly inform the design of the FSP Lunar Reactor-1, the first fission power plant intended for the Artemis Base Camp by 2030. NASA is developing the two reactors through a mix of in-house design, partnerships with other federal agencies, and contracts with major aerospace and nuclear firms. NASA serves as the prime integrator, with DOE providing expertise in nuclear fuel and reactor safety.
Lanteris Space Systems is building the PPE, which was originally part of the Gateway programme and is being repurposed as the spacecraft bus for SR-1 Freedom. Lockheed Martin, a primary contractor for the FSP project, is developing one of three prototype designs for the lunar reactor to meet the goal of 40 kW of continuous lunar power by 2030. While all three must use HALEU fuel and fit within a 6-tonne mass limit, they differ in their cooling and power conversion methods.
Westinghouse is adapting its terrestrial eVinci microreactor for space, focusing on extreme simplicity and a lack of moving parts. It uses a solid “matrix” core with passive heat pipes to pull heat away from the reactor eliminating the need for complex pumps or liquid coolants that could leak or freeze during the lunar night. It is paired with Stirling engines to convert heat into electricity.
Lockheed Martin’s design emphasises a modular architecture that can be scaled from a small 5-10 kWe unit to 100 kWe for industrial lunar activity. It uses gas-cooled technology and relies on dynamic Brayton cycle power converters. While more complex than heat pipes, Brayton cycles are generally more efficient at higher power levels, which supports Lockheed’s vision of scaling for future Mars missions.
IX (a joint venture of Intuitive Machines and X-energy) uses TRISO-X fuel which can withstand temperatures four times higher than conventional nuclear fuel without melting. The team includes Boeing and Maxar, leveraging their experience with International Space Station (ISS) power management and high-power satellite thermal rejection. Like Westinghouse, they are looking at Stirling-based conversion systems.
NASA is developing a modular lunar power architecture, LunaGrid, to integrate nuclear reactors with the Artemis Base Camp. This will allow multiple power sources, including nuclear fission and solar arrays, to share energy across various surface assets. To safely distribute the 40-100 kW generated by nuclear reactors, NASA is transitioning to a high-voltage primary distribution system, which significantly reduces the mass of the cables required to carry power over long distances.
For radiation protection, nuclear reactors will be positioned 1-3 km away from the crewed habitat, often placed over the lunar horizon and power will be transmitted by specialised high-voltage cable and spool systems.
Because different lunar assets (rovers, habitats, science suites) often operate at different voltages, NASA is developing the Universal Modular Interface Converter (UMIC), which acts as a “universal plug,” converting the high-voltage grid power into the specific voltage needs. This ensures that both government and commercial equipment can connect to the same nuclear-backed power source.
NASA is collaborating with companies like Astrobotic to test these distribution technologies early. The LunaGrid-Lite project, targeted for a 2026 test flight, will demonstrate the first robotic unreeling of a 1 km high-voltage cable on the Moon using a CubeRover. The aim is to transmit 1 kW from the lander to the rover to verify performance in one-sixth gravity and assess power losses due to the lunar regolith (fine surface layer).
NASA’s timeline for testing the LunaGrid and its associated power distribution technologies is structured through a series of robotic Commercial Lunar Payload Services (CLPS) missions that serve as precursors to the 2028 crewed landing.
In phase 2 (2027–2028), NASA aims to fast-track the delivery of lunar terrain vehicles (LTV) through CLPS. These vehicles are expected to be the first “mobile nodes” on the grid, capable of both drawing from and potentially contributing power to the network. UMIC will be tested to ensure the grid can handle various power loads from different international and commercial partners.
Phase 3, Operational Grid & Nuclear Integration (2028–2030) will see transition to a fully functional, reliable utility service alongside the first crewed landings. Astrobotic aims to have the first commercial grid service active at the South Pole by early 2028, coinciding with the Artemis IV landing mission. The final step in the current roadmap is the integration of the 40 kWe nuclear reactor, which will become the primary “always-on” power source for the grid, supporting continuous human habitation through the lunar night.
To navigate the treacherous terrain of the lunar South Pole and lay cables for LunaGrid-Lite, NASA and its partners are using a combination of specialised scouting and semi-autonomous driving. NASA is using data from the ShadowCam instrument (on Korea’s Danuri orbiter) to create high-resolution maps of permanently shadowed region. Because sunlight is unreliable, the CubeRover (the primary cable-layer) will rely on LiDAR (Light Detection and Ranging) to create a 3D map of the ground in total darkness.
Laying a cable in 1/6th gravity is physically difficult – a cable that is too loose can drift and snag, while one that is too tight can flip a lightweight rover. The spool on the Griffin-1 lander is equipped with active braking and tensioning, feeding the cable out at the required speed to ensure it lays flat against the regolith. The cables are designed as ultra-thin ribbons enabling them settle into the lunar dust.
As the lunar surface temperature can swing from 120°C to -230°C, the cable is insulated with specialised polymers that remain flexible at cryogenic temperatures. Designers must also ensure the high-voltage electricity running through the cable will not overheat it.
According to Science News, NASA’s announcements come at a troubling time for the agency. “Last May, the Trump administration proposed slashing NASA’s science funding for fiscal year 2026 in half, though Congress ultimately rejected the cuts. With the president’s fiscal year 2027 budget request looming, it’s unclear how much financial support NASA will have to achieve its goals.”
NASA also lost a large amount of personnel and expertise in 2025 under the administration’s efforts to shrink the federal government. “The agency just had its largest loss, percentage wise, of its workforce in a single year,” said Casey Dreier, the chief of space policy at the Planetary Society, headquartered in Pasadena.
The announcements came about a week ahead of the scheduled launch of Artemis II, originally planned for early February but delayed by leaks, and about a month after NASA cancelled its 2027 moon landing and scheduled two more for later on. “On the face of it, this is very exciting,” noted planetary scientist Paul Byrne of Washington University in St Louis. But for any of NASA’s plans to become reality will take a lot of money. “The history of human and robotic spaceflight is littered with ambitious timelines that are never fully realised,” he said.
For NASA’s nuclear dreams to be realised, it will have to get to work and secure funding. “If they’re serious about doing a nuclear-powered lunar base they’re going to have to start working with international partners and industry soon,” Byrne emphasised. “Within the next six to 12 months we’ll have to see positive indications that NASA’s budget will not just stay stable but grow.”
He added: “Honestly, it’s a wait-and-see. We are at an inflection point. This could either be a damp squib, like we’ve seen so many times before. Or, in hindsight, it could be the time we look back at and say, ‘This is when things began to turn around’.”
