The space station that NASA has spent years designing to orbit the Moon may never serve that purpose. Instead, the agency is now exploring a plan to gut its Gateway lunar station, bolt on a nuclear thermal propulsion system, and hurl the whole thing toward Mars — a move that would represent one of the most dramatic pivots in the history of American human spaceflight.
The concept, revealed in a detailed report by Ars Technica, emerged from internal NASA studies examining how existing hardware investments could be repurposed for a crewed Mars mission. Rather than building an entirely new deep-space transport vehicle from scratch — a prospect that would cost tens of billions of dollars and take decades — the agency is asking a simpler question: What if we already have most of what we need?
Gateway, as originally conceived, was meant to be a small space station in lunar orbit. A waypoint. Astronauts aboard Orion capsules would dock there before descending to the lunar surface in SpaceX’s Starship Human Landing System or other vehicles. Its core modules — the Power and Propulsion Element built by Maxar Technologies and the Habitation and Logistics Outpost, or HALO, built by Northrop Grumman — are already deep into development. The PPE and HALO modules are expected to launch together as early as 2027 or 2028 aboard a SpaceX Falcon Heavy rocket.
But the Artemis program has been under sustained political and budgetary pressure. And with the Trump administration signaling a desire to accelerate Mars timelines while questioning the value of a permanent lunar orbital presence, NASA engineers began studying alternatives. The result is a plan that reads less like traditional aerospace program management and more like a hot-rod builder eyeing a sedan in a junkyard.
Here’s the basic architecture. Take Gateway’s core modules. Strip away components designed exclusively for lunar operations. Add additional habitation volume — possibly an inflatable module from Sierra Space or a similar provider. Integrate life support systems capable of sustaining a crew for the roughly nine-month transit to Mars. Then attach a nuclear thermal propulsion stage to the whole assembly.
Nuclear thermal propulsion, or NTP, is the linchpin. Unlike conventional chemical rockets, which burn propellant in a combustion chamber, an NTP engine passes hydrogen propellant through a nuclear reactor core, superheating it and expelling it through a nozzle at extremely high velocities. The result is roughly twice the specific impulse of the best chemical engines — meaning far more thrust per pound of propellant. NASA and DARPA have been jointly developing NTP technology under the DRACO program (Demonstration Rocket for Agile Cislunar Operations), with Lockheed Martin as the prime contractor and BWX Technologies supplying the reactor. A flight demonstration is currently targeted for the late 2020s.
The math matters here. A crewed Mars mission using chemical propulsion alone would require enormous quantities of fuel, massive spacecraft, and transit times that push the boundaries of what human bodies and life support systems can endure. NTP changes the equation. It doesn’t eliminate the challenges, but it compresses transit times and reduces the propellant mass required, making the whole endeavor more feasible with hardware that actually exists or is close to existing.
According to the Ars Technica report, the NASA study envisions a phased approach. First, Gateway’s PPE and HALO modules would launch and potentially conduct limited operations in lunar orbit — or skip that step entirely. Then additional modules would be launched and assembled, either in lunar orbit or in a high Earth orbit. The NTP stage would be launched separately and mated to the growing stack. Once fully assembled and checked out, the vehicle would depart for Mars with a crew aboard.
The crew size under discussion is small. Four astronauts, possibly fewer. The habitable volume would be tight but survivable, especially if an expandable module is included. Life support would need to be largely closed-loop, recycling water and air with minimal resupply — technology that NASA has been testing aboard the International Space Station for years, though never at the duration or distance a Mars transit would demand.
There are enormous open questions. Radiation shielding is one. Beyond Earth’s magnetosphere, astronauts would be exposed to galactic cosmic rays and the risk of solar particle events for months on end. Gateway’s original design included some radiation protection, but nowhere near what a Mars transit would require. The NTP reactor itself adds another radiation source, though the plan calls for placing it at the far end of the vehicle stack with shielding between the reactor and the crew quarters.
Then there’s the matter of what happens when the crew arrives at Mars. The current concept, as described, focuses primarily on the transit vehicle — getting humans from the Earth-Moon system to Mars orbit. Landing on the Martian surface is a separate and arguably even harder problem. NASA has studied Mars landers for decades without converging on a single design. SpaceX’s Starship is one candidate, but landing a Starship on Mars with humans aboard involves entry through a thin atmosphere, supersonic retropropulsion, and surface operations in an environment where a rescue mission is physically impossible for years.
So the Gateway-to-Mars concept is, in a sense, solving only half the problem. But it’s the half that has historically received the least attention. NASA has spent decades studying Mars surface architectures while largely hand-waving the transit vehicle. This plan flips that priority.
The political context can’t be ignored. The Trump administration has made Mars a stated priority, with NASA Administrator Jared Isaacman — the billionaire astronaut who commanded SpaceX’s Polaris Dawn mission — publicly pushing for accelerated timelines. Isaacman has expressed skepticism about Gateway’s original lunar mission, questioning whether a station in lunar orbit is necessary when SpaceX’s Starship could potentially fly directly to the lunar surface from Earth. Repurposing Gateway for Mars would resolve that tension neatly: the hardware investment isn’t wasted, but it serves a destination the administration actually cares about.
Congress, however, has its own views. Gateway’s modules are being built by major aerospace contractors with facilities and jobs spread across multiple states. Northrop Grumman’s HALO module work is centered in Arizona. Maxar’s PPE work involves facilities in California. Changing Gateway’s mission doesn’t necessarily threaten those contracts — the hardware still gets built — but it does change the timeline, the integration requirements, and the testing regime. Congressional appropriators tend to get nervous when programs shift underneath them.
There’s also the international dimension. Gateway was designed as a multinational project. The European Space Agency is building the ESPRIT refueling module and the International Habitation Module. The Japan Aerospace Exploration Agency is contributing life support and thermal control systems. The Canadian Space Agency is providing Canadarm3, a robotic arm. These partners signed on for a lunar station. Redirecting Gateway to Mars would require renegotiating those agreements — or potentially leaving some partners behind.
ESA officials have been cautious in public comments. A lunar Gateway served European strategic interests by giving ESA astronauts access to the Moon without building their own heavy-lift rocket or lander. A Mars mission is more ambitious but also more uncertain, and European space budgets are under pressure. Whether ESA would follow NASA to Mars or peel off to pursue its own lunar ambitions remains an open question.
The technical heritage of nuclear thermal propulsion adds another layer of complexity. The United States successfully tested NTP engines during Project NERVA in the 1960s and early 1970s, demonstrating that the technology works. But NERVA was canceled in 1973, and no NTP engine has flown in space. The DRACO program aims to change that, but it’s still in development. Flying a nuclear reactor in space raises regulatory, safety, and public perception issues that chemical propulsion simply doesn’t. Launch failures involving nuclear material — however unlikely — would be politically catastrophic.
NASA’s study reportedly addresses this by proposing that the NTP stage be launched with the reactor in a cold, subcritical state. The reactor wouldn’t be activated until the vehicle is safely in a high orbit, well away from Earth. This is the same approach used for radioisotope thermoelectric generators on missions like Curiosity and Perseverance, though an NTP reactor is orders of magnitude more powerful than an RTG.
The timeline under discussion is aggressive by NASA standards but glacial by SpaceX standards. Internal documents suggest a possible crewed Mars flyby — not a landing — in the mid-to-late 2030s, with an orbital mission following in the early 2040s. A surface mission would come later still. Elon Musk has talked about sending humans to Mars by 2030, a date virtually no one outside SpaceX considers realistic. But the pressure Musk exerts on the conversation is real. NASA’s plan can be read partly as an attempt to present a credible alternative that uses government hardware and maintains the agency’s central role in deep-space exploration.
And there’s a deeper strategic calculation. If SpaceX does develop the capability to send Starship to Mars independently — with or without NASA — the agency risks becoming irrelevant to the most ambitious goal in human spaceflight. By repurposing Gateway, NASA positions itself as the integrator of a Mars transport system, the entity that brings together propulsion, habitation, life support, and international partnerships into a coherent architecture. That’s a role no private company can easily fill, at least not yet.
The engineering challenges remain formidable. Closed-loop life support for a 500-plus-day mission (transit to Mars, time in orbit, transit home) has never been demonstrated. The psychological effects of confining a small crew in a limited volume for that duration, with no possibility of abort or rescue, are poorly understood. Communication delays of up to 24 minutes each way mean the crew would need to operate with far greater autonomy than any previous space mission.
Food is a problem nobody talks about enough. Current space food has a shelf life of roughly 18 to 24 months. A Mars mission would push that to the limit. NASA has been researching food production in space — growing crops under LED lighting in microgravity — but these systems are experimental and nowhere near capable of feeding a crew.
Medical care is another gap. On the ISS, a seriously ill or injured astronaut can be back on Earth within hours via a Soyuz or Crew Dragon capsule. On a Mars transit, the crew is on its own. The vehicle would need to carry surgical supplies, pharmaceutical stocks, and diagnostic equipment, and at least one crew member would need significant medical training. NASA has studied telemedicine for deep-space missions, but the communication delay makes real-time consultation impossible.
Despite all of this, the Gateway-to-Mars concept has a certain elegant pragmatism. It doesn’t require inventing entirely new spacecraft. It repurposes hardware already under contract. It builds on NTP technology that the United States has already demonstrated at the component level. And it gives NASA a credible, concrete plan for Mars that doesn’t depend on any single commercial provider.
Whether it survives contact with congressional budgets, international partners, and the next presidential administration is another matter entirely. NASA’s history is littered with Mars architectures that looked promising on paper and died in the appropriations process. The Space Exploration Initiative of 1989. The Design Reference Architecture studies of the 2000s. The Journey to Mars branding of the Obama era. Each promised a path to the Red Planet. None delivered.
But this time, something is different. The hardware is real. The PPE and HALO modules are being built right now, in factories, by engineers who expect to see them fly. The DRACO nuclear engine is under active development with flight hardware in production. SpaceX’s competitive pressure creates a political urgency that previous Mars plans lacked. And the current administration, whatever its other priorities, has made Mars a stated goal with a NASA administrator who has personally flown in space and has little patience for bureaucratic delay.
None of that guarantees success. The gap between a study and a mission is vast, filled with technical reviews, budget fights, and the relentless entropy of federal program management. But for the first time in decades, NASA has a Mars plan that starts with hardware that exists rather than hardware that might someday be funded. That alone makes it worth watching.
The next major decision point comes later this year, when NASA is expected to formally announce whether Gateway will proceed with its original lunar mission, pivot to Mars, or attempt some hybrid approach. Industry sources suggest the announcement could come as early as the fall, timed to the fiscal year 2027 budget request. Until then, the engineers keep building, the reactors keep getting tested, and the question that has haunted American spaceflight for half a century — when do we go to Mars? — inches closer to having an actual answer.
WebProNews is an iEntry Publication