Salida / Great Sand Dunes, CO & Mars College, Salton Sea — ~$700 machine,
~100% on-site materials, zero added water, zero tensile elements. Mars is the endpoint; the gatehouse is the qualification prototype.
For Mars College: the pitch in one line — a $500 log splitter becomes a brick press,
local dirt becomes a storm-sheltered vaulted structure in ~10 days, with zero water and zero
purchased building materials — and every step of the method is a rehearsal for building on Mars.
What we'd love feedback on: the Salton Sea deployment site, soil suitability, who wants to
help press bricks, and whether the desert campus wants a unit of its own.
Everything below is the honest current state — designs and numbers done, first purchase not yet made.
Where it stands
Design
DONE
Parametric OpenSCAD model, manifold-clean STLs, enclosure ray-tested. Two cells kept.
Structural screening
DONE*
Kern/foothold screening passes. *Preliminary calcs, not engineering certification.
Procurement
IN PROGRESS
All parts located on Craigslist — nothing purchased yet.
Press build
NOT STARTED
Next: buy run → weld mold box → gauge + fluids.
Soil testing
NOT STARTED
Jar test, squeeze-ball moisture, crush gate decide site viability.
Structure build
NOT STARTED
~9–11 days all-in solo once pressing starts.
The machine — ~$500 core
A stock towable 25-ton log splitter IS the press: beam = frame, cylinder = ram,
engine + two-stage pump + valve + reservoir = the power pack. One fabricated part —
the mold box — bolts on with zero alteration (unbolts back to a working log splitter).
8×12×4″ jumbo bricks at ~520 psi face pressure; standard bricks for vault bands.
Extras in the design: echea / passive-bass acoustic cavities in the ribs, woofer discs,
storm threshold, vent shutters, paver cap armor. Amenities spec drafted: stove flue,
timber door leaf, paver floor, bed platform (~$270 purchased).
Teleop earthwork — Chicago Electric 3,000 lb winch w/ wireless remote (~$150, found on Craigslist) + slip pan + deadman anchor: the remote is the teleop. Same winch remounts on a monorail as the brick hoist.
Press auto-cycle — wiper motor + cam + limit switches on the valve lever, ~$30–80 scrounged.
Brick-placement gantry — Cartesian rail over the wall, ~$1–3k DIY. Our leaned-vault method makes this the simplest published-compatible robot task: every course rests on the party wall, no support robot needed (research systems like VaultBot / NASA ARMADAS need cooperative support arms — ours don't, and cost ~1%).
Loaders/tractors — demoted to "if a deal lands": Ford 2N $1,000 Avondale, Int'l+loader $2,500 Gardner, JD 1050 $3,500 Pueblo (stale = negotiable).
Mars transfer — verified against literature
Dry-compression is a published route — Martian simulant compacts into blocks with no binder/water (Nature Sci. Reports). Our ~520 psi is below published test pressures → MICP is the binder fallback (3.3–5.65 MPa on simulant, survives 1% perchlorate).
Gravity: 0.38g scales thrust and confining weight together — Earth is the harder case.
Shell ≠ habitat: dry-stack shell is a radiation shield around a pressurized bladder — shielding claims need >1 m regolith (dose peaks at 0.4 m).
Teleop ≠ Mars ops: 4–24 min light delay → supervised autonomy is the real architecture; the Earth winch-demo proves the task is remote-buildable.
Carbon: Earth envelope can take 4–6% coarse biochar (~8–10 t CO₂e buried/structure) — functional (moisture + acoustic fill), sequestration rides along. Mars substitutes scoria.
Honest limits: structural numbers are in-house screening, not code-certified ratings —
no Airbnb/occupancy use without licensed structural/geotech/life-safety review and permits.
Craigslist listings are unverified until inspected (tonnage, cylinder condition, hose/valve integrity).
The winch is a demonstrator-scale earthmover, not a production machine.