Division 04 · Earth Analog Program
You cannot iterate on Mars. Every question worth asking — regolith chemistry, closed-loop water, spectrum, pressure — has to be answered on Earth first, inside facilities built to lie convincingly about being somewhere else.
The Gap
A simulator is a model of what we believe. An analog facility is where that belief gets tested against something that can actually kill a plant. The distance between the two is the entire near-term research programme.
So the Earth analog site is not a marketing prop. It is the thing that converts RPA from a brand with a thesis into a company with data — and it is what Founding Plot packages are funding.
Facility Path
Each stage costs more and proves more. We climb only when the previous stage has produced something worth carrying up.
Leased or partnered bench space inside an established CEA programme. This is where the science that actually matters happens: LED spectrum recipes, hydroponic nutrient work, closed-loop water recovery, and perchlorate handling in regolith simulant. No airlocks, no crew, no theatre — just repeatable experiments at a cost a young company can carry.
A hermetically sealed, pressure-controlled habitat with greenhouse heritage — the closest thing on Earth to growing inside a Martian dome. This is where reduced pressure stops being a spreadsheet number and starts being a structural and biological constraint. It is also, not incidentally, where the footage gets genuinely compelling.
Short crop campaigns run under full mission constraints — limited water, limited power, crew in suits, resupply on a schedule. Desert research stations already operate greenhouse modules on exactly this basis. What this stage proves is not botany but operations: whether the system survives contact with the way people actually work off-world.
Arctic impact-crater sites are the highest-fidelity Mars analogs on Earth, with real greenhouse research heritage. They are also logistically brutal and expensive, and the science they add over Stage 2 is narrow. We name it here for completeness and to be clear that we are not starting there — anyone who tells you they are, at our stage, is selling something.
How We Choose
The most Mars-like place on Earth is the wrong first choice if you can only afford to go once.
How many grow cycles per year? A vine that fruits once annually punishes a site you can only reach in summer.
Bench space in a university lab runs a fraction of a crewed rotation. Early on, volume of data beats prestige of location.
Existing sensing and data infrastructure is worth more than raw realism. Unmeasured results are anecdotes.
A sealed habitat films better than a grow rack. Media value is a legitimate tiebreaker for a company whose flagship is a brand.
Grow Programme
Vines are slow, expensive, and unforgiving. They are the last thing you put into a loop you have not yet proven.
Lettuce and radish. Weeks, not seasons. Their only job is to prove the water loop, the nutrient line, and the sensors work.
The other half of wine, and the cheapest thing to fly. Low mass, dormant-stable, and a credible early payload candidate.
Dormant cuttings into treated regolith simulant. Survival and root establishment first; fruit is not the question yet.
Full spectrum recipe, reduced pressure, closed loop. The first grapes grown under Martian conditions on Earth.
No facility agreement is in place. The sites described above are categories we are evaluating, not partners, sponsors, or affiliates of RPA. None of the named institutions have endorsed, reviewed, or agreed to anything, and nothing here should be read as implying otherwise.
No grow campaign has run yet. Every result, timeline, and capability on this page is a plan. Member benefits tied to the analog site — plaque credits, site visits — are fulfilled once a facility is actually secured, and we will say publicly when that happens.
If you operate a CEA lab, a sealed habitat, or an analog station and this programme is interesting to you, that conversation moves faster than anything else on our roadmap.