Thirteen mechanisms.
One grid, one port.

1. Facet panels are glass-FRP sandwiches with a PIR core: 80-120 mm for temperate sites; polar sites need 200-250 mm PIR-equivalent (U about 0.09-0.11 W/m2K at panel centre). A mycelium core is an option only where mass and U targets allow (it adds 7-24 kg/m2 at 120 mm).
2. Every panel edge carries a structural closeout of pultruded GFRP or hardwood, 15-25 mm deep. This closeout, not the core, is the bearing surface; a 10-20 mm aerogel strip sits behind it to keep the thermal line continuous.
3. Each test article uses one thickness method. The Miura wall uses Tachi tapered panels, which have a limited fold range, so its folded thickness is taken from the fold geometry. The vault uses Chen/You offset panels, which fold flat but need stack depth for the offset hinge lines. No packing figure assumes both at once.
4. Rigid-foldability is checked before tooling: the crease pattern is run in a kinematic simulator (MERLIN or Origami Simulator) and must show zero facet strain through the whole fold. Patterns that need facet bending or snap-through (for example the classic Yoshimura) are not used for rigid thick panels.
5. At each crease the rotation axis sits on one face, and that face alternates between mountain and valley creases. A tight vacuum-laminated PVDF-coated polyester membrane spans the crease on the axis face. On the opposite face the edge gap changes by about t x dTheta, so the membrane there is a slack loop or bellows of at least 1.2 x t x dTheta (about 230 mm for t = 120 mm and a 90 degree fold change; about 150 mm for t = 80 mm), or a bonded EPDM or silicone strip with at least 300% elongation.
6. The inner layer is the vapour control layer and the outer layer the rain screen; at each valley the space between them drains to the sill gutter, with a heat-traced or self-draining outlet at cold sites.
7. The offset or taper geometry lets each crease close only to its design dihedral; faces meet and stop it, so a wall cannot over-rotate even if a winch is mis-operated.
8. At the design dihedral the edge closeouts bear on each other in compression. Under wind reversal the tension side of each crease moment passes through the latch, so the latch is a primary structural member: latch tension T is about M_crease / (0.8 t), and it is sized for fatigue as well as preload.
9. The latch is thermally broken: a GFRP or polyamide body with a bronze cam only at the bearing point, mounted on the inside face behind the vapour layer so it does not pierce the membrane.
10. An EPDM bulb gasket inside the membrane is compressed by the same latch preload as a second air barrier.
11. Each latch carries a Hall-effect position sensor and a strain gauge; the module passport records engaged or not, and TSURU verifies every latch after deployment.
12. A failed membrane or bellows strip is cut out over 300 mm and re-laminated with a field vacuum kit as an obi patch (target 30 minutes).

Risks UV and abrasion ageing of the coated fabric and bellows at outside creases · Ice formation in valley drainage channels and condensation at crease lines in cold climates · Fire performance of FRP and foam cores (no code route proven yet) · Thermal bridging at latches and edge closeouts · Latch fatigue under wind reversal · PFAS regulation of fluoropolymer coatings

EXPERIMENTAL: each ingredient is a prototype or lab result; the weathertight, insulated building hinge with an opening-face bellows is unbuilt and is the single most important risk in ORIJO.

1. ASHI legs are set and levelled; MUTOIS floor cassettes unfold onto them to form a 3.64 x 7.28 m floor diaphragm with a sill beam on each long side.
2. The shell ships as two half-length (3.64 m, 2-bay) packs. Each is lowered onto two kinematic datum pins, at the fixed gable for the first half and at the splice frame for the second; transit straps are released by app command after a pre-check.
3. The crease pattern is a rigid-foldable quadrilateral-panel (Miura-derived) cylinder with degree-4 vertices, built with Chen/You offset panels and checked in a kinematic simulator at zero facet strain. It has one degree of freedom, so the whole half-shell moves together; there is no crown-first transition front.
4. The moving gable portal is winched along floor rails. Because span and rise change as the cylinder extends, the haunch shoes run in PTFE-lined slides along the sill rails; the slide travel is designed, not left to friction. Near the folded state the pull force rises steeply, so the first part of the stroke uses a geared winch or a jack start, and counterweights (optional) balance the rising facets.
5. If a staged opening is needed, it comes from staged latches or control cables, not from the pattern.
6. At full extension the one-way offset stops engage; the two halves are joined at a latched splice ring frame; TSURU walks the ridge rail and closes each latch at 910 mm pitch, logging sensor state.
7. Haunch facets lock into the sill shoe; tension ties across the span resist spreading thrust.
8. Kirigami shoji gables (F6) unfold by one string; the door cassette and the KURA core dock by JO-PORT.
9. A blower-door fan in the KURA core pressurises the shell; leaks are localised by TSURU thermal camera and fixed before handover.
10. Load path: snow and self-weight travel as folded-plate shell action along ridges and valleys to haunch shoes, sill beams, ASHI legs and anchors; wind uplift returns through edge-closeout bearing, latches in tension and base ties. Retraction is the reverse.

Risks Snow drift on crease valleys creating unbalanced loads · Tolerance stack-up across about 300 m of crease and dozens of vertices · Haunch slide friction and peak winch force near the folded state · Pack envelope not yet confirmed by fold simulation · Code approval of a novel folded shell · Hinge wear during repeated relocations

EXPERIMENTAL: rigid-foldable thick-panel cylinders exist as models and small prototypes; an insulated, airtight, snow-loaded occupied vault is unbuilt.

1. Site scan by drone photogrammetry and a probe rod gives slope, soil class and obstacles; the configurator selects leg count, pile length and cut-off heights.
2. Coarse slope is taken up by pile cut-off height or a stepped sill. Each telescoping leg then carries a jack for fine trim only: +/-115 to +/-230 mm (TenFold spec: 9 in total surface variability, hydraulic), which is about 3% (1.8 deg) over 7.28 m.
3. Screw piles are driven by a compact excavator auger head (target 15 min per pile). No trench is dug.
4. Lateral load goes through X-bracing between legs or a moment-rigid sill-to-leg connection. Each leg and pile head is rated for at least 1.5 x 1/n of the wind shear (about 5-10 kN per leg for 4-8 legs).
5. Height is held by a mechanical pin or threaded lock nut with a published load rating. A separate replaceable impact absorber is fitted only where an impact case exists. Snow build-up and settlement are handled by re-jacking, as at Halley.
6. Leg capacity comes from commercial jacks, e.g. 2-5 t rated per leg against about 5.4 kN static per leg for YOSHI-4 on 4 legs, plus wind and snow combinations.
7. HRV intake air is protected from frost without a trench: a glycol loop in thermal helical piles feeds a pre-heat coil, or an electric frost pre-heater (about 0.3-1 kW) sits on the HRV. A separate earth-tube trench is an option only for multi-year sites.
8. Rooftops: zero penetrations only for HOSHI-class pods, subject to a structural check (even HOSHI-2 needs about 1-2 t of ballast). YOSHI-4 on an existing roof needs a spreader steel frame bearing on columns or beams, or mechanical tie-downs.
9. To leave, legs retract, piles unscrew and the site is restored.

Risks Frost heave on piles · Low lateral capacity of helical piles in soft soil · Wind sliding on roofs and roof structural capacity under ballast · Trench safety and excavation cost where an earth tube is chosen · Biofouling, fitting corrosion, permitting of open-water heat exchange, mooring and stability for the pontoon foot · Custom crate tooling (about USD 50-150k mould) if a jo crate is ever made

COMMERCIAL components (screw piles, jacks, bracing); EXPERIMENTAL as a levelled, braced leg system; SPECULATIVE for ski, pontoon, rock-plate, rooftop-ballast and regolith feet

1. The datum is spread over the whole module: two hardened steel balls (>= 58 HRC) at the port end and one at the far end, running in hardened steel V-block inserts pressed into the tile. The balls carry lateral location under light preload; the module's weight goes to flat bearing pads, so contact stress stays within the insert's capacity (quasi-kinematic).
2. The port box at one short end floats about +/-3 mm on a compliant mount, so the ports mate relative to the module, not the tile. It holds separate islands for DC 48-60 V, potable in, grey out, optional air duct and data (CAN/Ethernet + NFC).
3. The organ is lowered by TSURU (when available) or rolled by two people; lead-in cones guide the V-grooves onto the balls and a spring clamp preloads the seat.
4. NFC reads the module passport (ID, revision, mass, ratings, ports, firmware); the controller compares it with the site configuration.
5. Enable sequence for each fluid island: mate flat-face dry-break couplers, pressurise the closed interstitial volume, run a pressure-decay test on the transducer, then open the downstream solenoid isolation valve. Then DC precharge, then power. Any failure leaves that island isolated.
6. Urine never passes a dry-break coupler: it leaves by a gravity drain with an air gap, or by a flushable quick-disconnect with an acid-flush cycle against struvite.
7. Neighbouring tiles lock edge-to-edge with LVL or steel keys at 1-2 mm clearance and a wedge or cam take-up; the 0.5 mm datum is local to each tile, so tolerances do not stack across tiles.
8. Removal reverses the sequence; residual fluid falls into the tile drip tray; the passport logs the event.
9. Critical organs (water, battery) get a twin port on the adjacent tile, which costs one free 1.66 m2 jo per critical organ, so a replacement is live before the old one is pulled (shikinen).

Risks Dirt and ice in couplers · Struvite and calcium scale fouling in urine lines · House DC load above about 2 kW erasing the 48 V DC advantage · Standardising before the catalogue is stable · DC arc-fault protection maturity · WOTA Unit availability outside Japan and pilot regions

EXPERIMENTAL integration of COMMERCIAL parts (connectors, DC hardware); the water organ is a stand-in, because the WOTA Unit is PILOT (pre-mass-production, AC only, no port interface)

1. SIZE: the configurator computes harvest = people x (demand x (1 - net recovery) + 2-3 L drinking and cooking), with demand 60-100 L/person/day (water-conserving) and 90-95% net recovery in desert conditions, giving about 5-13 L/person/day. Yield per kg is taken from single-cycle passive data, not lab multi-cycle data.
2. CAPTURE (night): rack louvres open and low-power DC fans pull air through the sorbent beds - CaCl2-hydrogel composite or MOF-303 bonded to folded copper lamellae (v2 cassette). The rack runs one cycle per day; shaded daytime desert air (5-15% RH) is below the MOF-303 uptake step, so there is no daytime capture.
3. RELEASE (morning, about 07:00-11:00): the beds are shuttered into a sealed chamber and heated by a glazed solar-thermal collector loop; PV-T backsheet heat is used only as preheat or for CaCl2 composites. The governing check is p_sat(T_cond) / p_sat(T_bed) <= 0.1 for MOF-303 (e.g. condenser 25 C and bed about 70 C) or <= 0.2-0.25 for CaCl2 (bed about 55-60 C at a 25 C condenser). A DC trim heater fills cloudy-morning gaps.
4. CONDENSE: vapour goes to a condenser backed by night-cooled thermal mass or ground-coupled water that is separate from the potable tank, with a radiative-cooling coating on its shaded face; condensate drips into a weighed sump. Condenser temperature and the residual sorbent loading at the end of each release are logged.
5. VERIFY: conductivity, turbidity and temperature are logged; water defaults to the grey loop until in specification.
6. PURIFY: activated-carbon polish, then a calcite/dolomite mineral cartridge with CO2 dosing from a food-grade cylinder, then UV-C LED at 265-280 nm, then a chlorine residual for storage. Lithium checks apply only if a LiCl sorbent is used.
7. STORE: food-grade bladder in the base cells of the KRES mast (3-6 m gravity head) with a radiative-cooled top.
8. DISTRIBUTE: gravity to JO-PORT potable islands; DC booster pump; UV ring at each tap as the final barrier.
9. FALLBACK: fog-mesh skin on coastal deserts and ridges, a commercial hydropanel or AWG, or a truck fill port; the harvester never carries life safety alone.

Risks MOF availability and price unknown · Collector area per person larger than the roof allows at low-yield sites · Dust fouling of sorbent beds · Salt leaching from composites · Condenser mass too warm after hot nights

EXPERIMENTAL (passive sorbents at pilot scale; WOTA Unit pre-mass-production; integration unbuilt). The recycle-first architecture with commercial make-up is the investable core.

1. SEPARATE at the source: urine-diverting dry toilet (no blackwater), grey drains (kitchen sink via a grease interceptor), kitchen organics bin, crop residue chute, HRV/dehumidifier condensate line.
2. GREY: the MIZU organ runs biology, membrane, RO, UV and chlorine; up to 97% is reused with rainwater make-up (vendor spec); make-up water is logged separately; RO reject goes to solar membrane distillation, ornamental irrigation or the holding tank.
3. YELLOW: urine is stabilised by storage or acidification so ammonium (NH4+) stays the dominant species, then passes a clinoptilolite exchange bed that captures NH4+ and K+ at 10-20 mg N/g. The bed is regenerated in place with a KCl or NaCl rinse, which produces a liquid NH4/K fertiliser; it is not a disposable pod. (Alternative: full nitrification to nitrate liquor used directly as hydroponic feed, the Aurin/VUNA route.)
4. Charged zeolite also feeds zeoponic seedling trays and substrate beds in HATA; surplus nitrogen goes to outdoor beds or to a fertiliser offtake, because one NFT rack uses only a few percent of household urine N.
5. BROWN: dried faeces from the ventilated chamber are composted thermophilically for non-food landscape, or incinerated or collected where regulation requires.
6. GREEN: kitchen and crop residue (about 1-2 kg/day for 4 people) feed a BSF larvae bin on a 10-14 day cycle; larvae feed fish or poultry, frass goes to outdoor beds.
7. AIR-MOISTURE: plant transpiration and breath are condensed by the dehumidifier and HRV and returned to the grey intake.
8. LUNG: sealed farm modules carry a bladder that absorbs pressure swings and receive the living unit's CO2-rich exhaust through a HEPA and activated-carbon stage, within a stated humidity budget (target about 1,000 ppm CO2 in the grow zone).
9. SODIUM: recirculating nutrient is bled off on a sodium set point to the holding tank (urine carries about 2-4 g Na per person per day).
10. MONITOR: EC, pH, ORP, turbidity and Na online; quarterly lab panels for pathogens and pharmaceuticals; anything out of spec goes to the sanitary holding tank or septic connection.

Risks Regulatory acceptance of urine-derived fertiliser varies by jurisdiction · Pharmaceutical residues · Sodium and salt build-up in recirculating nutrient · Nitrogen surplus with no local outlet · Humidity, VOC and bioaerosol transfer from living unit to farm · BSF economics failed industry-wide (internal service only)

PILOT grey loop (WOTA Unit pre-mass-production; WOTA BOX commercial for showers); EXPERIMENTAL urine-to-zeolite and closed nitrogen loop

1. PLAN: the twin takes site climate, solar resource, people and diet, then returns crop mix (greens, herbs, cherry tomatoes or strawberries, duckweed protein), cassette count, kWh/day and kg/week.
2. SOW: a seed-metering cartridge drops pelleted seed into plugs in half-jo trays (910 x 455 mm) on mineral substrate at the nursery tier. Zeoponic (urine-charged zeolite) substrate is used for edible crops only after its separate bench trial passes pathogen and pharmaceutical assays.
3. MOVE: trays ride roller lanes; one linear shuttle per HATA cassette indexes trays between four tiers and re-spaces them onto wider-pitch trays at about day 10 and day 20 (Spread logic).
4. MONITOR: fixed cameras per tier measure canopy area, colour and tipburn; the twin updates a probabilistic harvest date per tray.
5. IRRIGATE AND FEED: NFT with mineral nutrient solution and automatic EC/pH dosing; transpiration recaptured.
6. LIGHT AND AIR: indoor tiers run DLI recipes at 14-17 mol/m2/day with fixtures of at least 2.7-3.0 umol/J; glazed YOSHI-6G vaults use daylight with LED top-up; occupant CO2 is ducted in through a filtered lung line.
7. POLLINATE AND HARVEST: cherry tomatoes sit in a rigid gutter grid; a rail arm pollinates by vibration or air pulse (shown for buzz-pollinated tomatoes) and picks by camera at known coordinates. Strawberries need even pollen spread, so a strawberry wall uses bumblebees or a brush fallback. Greens are harvested by whole tray at a cutting station.
8. CLEAN AND RECYCLE: trays pass a clean-in-place station; residues go to BSF.
9. MAINTAIN AND QUARANTINE: pump, LED and seeder cartridges are swapped through JO-PORTs; any cassette can be sealed and sanitised alone (Bowery pathogen lesson).
10. ADAPT: desert adds HIMURO cold store and shading; arctic runs fully sealed with LED heat recovered; a later outdoor phase adds FarmBot beds and a cable robot under BYOBU-PV fences.

Risks Energy cost of indoor greens (only high-value or remote-premium crops pay) · Crop pathogens in closed systems · Pick success below human level in cluttered canopies · Food-safety and GAP certification of urine-derived substrate · Vertical-farm sector failures colour investor perception

COMMERCIAL components; EXPERIMENTAL integration and autonomy level

1. ENVELOPE FIRST: the configurator picks panel class, kotatsu table, heated sleeping platform and HRV before any PV is sized, and plans a winter deficit at high latitude rather than a surplus.
2. GENERATE: MIURA wings over the engawa (about 4.5 kWp after facet, hinge and frame packing losses) and BYOBU vertical bifacial fences (snow-shedding, albedo gain); micro-hydro or small HAWT only as measured-site bonuses.
3. DISTRIBUTE: MPPTs feed a 48-60 V bus running in the floor cassettes to JO-PORT DC islands; point-of-load converters replace inverter stages; heavy loads stay within the house DC budget or move to 380 V DC or AC.
4. STORE ELECTRICITY: LFP inside the heated envelope for daily cycling; a 2-5 kWh sodium-ion bank in the unheated vestibule carries keep-alive loads (comms, sensors) down to its rated discharge floor so a heat-loss event never bricks the system.
5. STORE HEAT (default): surplus PV runs a CO2 heat pump into a 300-500 L insulated water tank, about 23 kWh thermal from about 8 kWh of PV at COP about 3.
6. STORE HEAT (optional hearth): a 0.4-0.6 t sand or soapstone radiant hearth, core at or below 250 C, with 150 mm microporous insulation and a rockwool jacket, about 25 kWh thermal, a 12-24 h buffer. It stands on a dedicated pier or slab pad rated at least 10 kPa, off the floor cassettes; touchable faces stay at or below 45 C (ISO 13732-1); a summer vent or bypass dumps heat outside; fire separation as for a stove.
7. HEAT AND COOL: CO2 heat pump makes 70 C water; HRV with frost pre-heat; radiative-cooled roofs, windcatcher and night-sky radiator cool hot sites.
8. PRIORITISE: firmware ladder - life safety (CO/CO2, smoke, comms) > water treatment > food (cold store, farm) > heat to people > comfort > guest extras - with forecast-driven load shifting.
9. OPERATE: Venus OS + Home Assistant run locally; MQTT telemetry over Starlink feeds the fleet dashboard; the site runs without the cloud.

Risks Surface temperature and fire safety of a hot hearth indoors · Floor loading of any thermal mass · Na-ion stationary ecosystem maturity and supply concentration · Winter PV deficit at high latitude

COMMERCIAL components (LFP, Victron, CO2 heat pumps, tanks); EXPERIMENTAL Na-ion stationary integration and the small radiant hearth

1. Every YOSHI ridge, the KURA spine and each farm bay carry a segmented T-slot rail. The rail sits on one panel and is broken at every transverse crease; datum sockets at 910 mm pitch are bolted through to the ridge beam, not to the skin, and have drained, self-cleaning cups.
2. TSURU is a 5-DoF bipedal inchworm: it locks one foot, swings the other to the next socket and locks again. Each foot is a 2-3 point tapered cone or petal lock with 20-30 mm passive lead-in (after the ARMADAS 3.0 cm petal capture), rated for at least 0.6 kN.m moment and 3 kN fall-arrest load, with lock-confirmation sensing.
3. Accuracy is local: socket-to-socket error is set by the socket and its lead-in, while hinge play and erection tolerance (several mm per joint) mean building coordinates come from a post-erection survey, not from the robot.
4. Load case: at full 910 mm swing the robot carries 5-8 kg; the 15 kg payload is handled only in double-stance mode with both feet locked. Joint actuators are sized for at least 400 N.m peak at ankle and hip (M = g x (m_body x L/2 + m_payload x L) x 1.7). Step time is 15-30 s, about 0.03-0.06 m/s.
5. A continuous fall-arrest trolley runs along every rail TSURU uses outdoors.
6. Tasks arrive as versioned contracts and are limited to ridge-adjacent work: latch close and verify, hinge thermal scan, filter/UV/sorbent swaps at JO-PORTs, seed-cartridge reload. PV and slope washing go to a separate brush crawler or to people.
7. TSURU is ferried by a person between modules; one robot serves one contiguous rail network.
8. Ground and cluster tasks use a rented legged inspector or a farm utility rover on scheduled service visits; cartridges of up to 65 kg can arrive routinely by drone at roadless sites.
9. Exceptions escalate to a remote operator over Starlink; anything outside the contract goes to a local human runner.
10. Renewal: sensor triggers or the shikinen schedule create a swap job; the new organ installs on the twin pad, commissions, and the old one leaves for the micro-factory.

Risks Ice, grit and water in sockets preventing lock · Liability for falls and for people below · Rail continuity across transverse creases · Scope creep toward general-purpose robots

EXPERIMENTAL (TRL 3: lattice-walking robots are indoor prototypes; crease-rail service robot unbuilt)

1. Each facet is a louvre on a central-axis pivot, with parallel pivot axes in each row so neighbours do not collide, mounted over the KASANE membrane with a ventilated gap. One face is BaSO4 white (solar absorptance about 0.03), the other carbon black (about 0.94); both faces are high-emissivity in the infrared, so the flip changes only daytime solar gain.
2. Heating facets have a thin 3-5 mm EPS/XPS stiffening core instead of aerogel, and winter dampers close the gap so it works as a solar air collector feeding the interior. Aerogel is used only on white-only cooling facets.
3. A commodity wax-cylinder opener or NiTi spring senses interior or gap air through a shaded sensing bulb, never the facet itself. A hysteresis band of at least 8 C (white above 26 C interior, black below 18 C) stops the facet's own state from driving it into a limit cycle.
4. The actuator does not lift the facet: about 70 N of wax force trips a pre-tensioned over-centre cam and toggle spring that rotate it 180 degrees. A detent or latch rated at least 500 N at the facet edge holds it against wind.
5. Baseline SKU: a seasonal manual flip twice a year (by hand, or by TSURU where a facet is ridge-adjacent). The self-flipping version is an option.
6. Farm vaults are ventilated by a wax-motor ridge vent sized at about 15-20% of floor area, with a cold lockout that keeps it shut below 5 C outside. Hygromorph flaps are only a secondary trickle vent.
7. In summer, the farm-vault glazing (not the opaque facets) gets a chalk-wash coat; at dry sites such as the Atacama a dry brush or minimal-water wash replaces rain washing.

Risks Soiling and reflectance loss in desert dust · Pivot and toggle fatigue · Wind flutter and back-driving · UV and creep of hygromorph flaps

TRL 2-3 integrated (lab cm-scale Janus membrane, hand-flipped; wax actuators TRL 9; self-flipping 1 m facet unbuilt)

1. Site pin: climate record, solar resource, humidity, snow and wind classes, and the local regulatory path (state modular or factory-built program, local appendices, RV/ANSI A119.5 for units on wheels) are looked up.
2. Typology and size: the customer chooses farm, hotel, home, research station or (for partners) lunar analog. US home SKUs are kept at or below 37.2 m2 by choice, to use the Appendix BB relaxations where a jurisdiction has adopted them; larger homes go through the standard IRC.
3. A rules engine resolves the order into jo tiles and organs from the fixed catalogue, as HAPPS does for Sekisui Heim.
4. Balance sheet: water (recycle first), energy (envelope first), food (DLI and m2), nights (ADR and occupancy) shown live.
5. Structural check: pre-validated FE templates per fold family and load class; occupied units are signed off by a partner engineering firm under an SLA (target 48 h), not a promise to the customer.
6. A binding kit and shipping price is issued with a capped site-works estimate (+/-15%) for foundation, utilities, access and permit fees; the customer pays a deposit, and the site price is finalised after a site survey.
7. The 15-working-day clock runs from deposit plus approved site survey to shipped kit; permit approval time is shown separately per jurisdiction.
8. Cut files, BOM and robot task contracts go to the nearest licensed micro-factory: a containerised kit (1-2 x 40 ft) deployed into a leased shed or tent of about 300-500 m2, certified under the state modular or factory-built program with third-party in-plant inspection; module passports are created.
9. At commissioning the site twin goes live on the fleet dashboard.

Risks Permit variance across jurisdictions · State modular certification per plant delaying occupied units · Partner micro-factory quality · Configurator promises exceeding physics or site reality

COMMERCIAL: kitting (HAPPS) and CNC kit libraries (WikiHouse); PILOT: licensed robotic micro-factories (AUAR); SPECULATIVE: origami SKUs and the physics-checked configurator

1. Each SKU has a published swap cycle (example targets: UV and filters 1 yr, sorbent cassettes 2-3 yr, farm cassettes 5 yr, membranes 10 yr, shell 20 yr).
2. Telemetry (water quality, energy, strain, latch state, crop health) feeds condition-based triggers.
3. Water organ swap: the new module is placed on the twin pad, connected to a twin-pad manifold with isolation and bypass valves and dry-break quick connects, flushed and disinfected through a dedicated port, and cut over only when an inline turbidity and chlorine sensor passes.
4. Battery organ swap: break-before-make through an automatic transfer switch; aged and new packs are never paralleled.
5. Returned modules go back by reverse logistics, are refurbished at the micro-factory, and are re-leased or resold; refurbishment yield is measured against the residual value assumed in the lease.
6. Lease price L = C x CRF(r,n) - R x SFF(r,n) + M + Lg, with CRF = r/(1-(1+r)^-n) and SFF = r/((1+r)^n - 1); C is module capex, R residual value, M maintenance, Lg logistics.
7. In emergencies, modules come from a dedicated surge pool, not from active leases unless the lessee opts in, under agreements set up with the municipality in advance.
8. Lease contracts include an open-firmware escrow so customer sites run if Anyground fails.

Risks Balance-sheet intensity of owning hardware (cost of capital dominates) · Reverse-logistics cost per swap at distributed sites · Too few failure events in a small pilot to estimate availability · Customer preference to own

COMMERCIAL model elements (product-as-a-service, maintenance-inclusive leasing, OpCo/PropCo); EXPERIMENTAL as a lease for small distributed habitat organs

1. A rigid domed (torispherical) pressure head with a hatch cut-out carries the KURA systems layout - ECLSS, water, power and the stowed softgoods bladder (X-Hab architecture). It is a new pressure vessel, not KURA box geometry.
2. The bladder is deployed and pressurised between two end-caps; a Vectran-class woven restraint layer carries hoop load, and longitudinal straps carry the axial end-cap load F = p x pi x r^2 into a ring frame at each head (about 714 kN at 101 kPa or 400-500 kN at 56-70 kPa for a 3 m diameter).
3. Softgoods are certified to NASA practice: burst at least 4x the maximum design pressure (e.g. 4 x 15.2 = 60.8 psi).
4. A YOSHI rigid-panel shell with lockable hinges deploys over it as a regolith-bag carrier and micrometeoroid/thermal layer, not as a pressure vessel; it is sized for about 3.9 kPa of lunar shielding load.
5. Robotic pads and berms are laser-sintered from regolith (ICON LVMT) before crew arrival.
6. A source-agnostic power brick (DC bus + hydronic loop) accepts a 10 kWe fission unit; on Earth the same brick is tested with PV and a basalt-gravel thermal store.
7. Sealed HATA cassettes and algae/duckweed shelves close part of the food and gas loop; ISS/Lunar Palace-class water closure needs urine processing (vapour-compression distillation plus brine processing).
8. Circadian lighting and fractal-patterned kumiko interiors address the 14-day night and isolation.

Risks No flight heritage; Earth analogs reach TRL 3-4 for layout and operations only · Pressure-vessel redesign and softgoods certification · Regulatory and launch dependence · Distraction from Earth revenue

SPECULATIVE