What we have built, and what we have not

As of this writing, we have not built the prototype house yet. We have designed the system. We have specified the materials. We have calculated the loads. We have planned the tests. We have identified the constraints. But the house itself is still on paper.

This is not a weakness. It is honesty. We are publishing our design before we build it because we believe the idea is worth sharing, and because we want the scrutiny of people who know more than we do. The prototype build in Algarve, Portugal, is scheduled for Q3 2026. When it is built, every claim on this site will be backed by data and a receipt. We will publish the results, the failures, and the adjustments, in real time, under the same open licence as the design itself.

Here is exactly what we will test, and the protocol we will follow.

What we do not know yet

These are the active engineering questions. We are not pretending they are solved. The prototype build in Algarve, Portugal (Q3 2026) is designed to answer every one of them.

→ Active

Lateral bracing

Corner panels and shear walls provide bracing under our assumptions. The OSB-faced panels create shear diaphragms. But the full diaphragm capacity of a pallet-core panel versus a solid-timber-core panel has not been tested. We assume the OSB skins carry the shear and the pallet blocks provide fastening points. This will be verified with full-scale racking tests at LNEC.

Mitigation: Temporary bracing during construction is standard practice. Permanent bracing may require additional timber diagonal straps in high-wind zones.

→ Active

Condensation risk

The assembly is vapour-open: OSB is vapour-permeable. Rock wool is vapour-permeable. Limewash is vapour-permeable. The entire wall breathes. But in a heating climate with high internal humidity, interstitial condensation at the OSB faces is a risk we have not quantified.

Mitigation: 12-month monitoring with humidity sensors in cavities. WUFI modelling before build. Breather membrane on cold side. No vapour barrier on warm side.

→ Active

Code approval pathway

Portuguese building code (Regulamento Geral das Edificacoes Urbanas) allows engineered timber systems with stamped structural calculations. But an RCP panel is not a standard product with an ETA (European Technical Assessment). We need a structural engineer to produce project-specific calculations and a test body to provide evidence.

Mitigation: LNEC (National Laboratory for Civil Engineering) testing planned for Q3 2026. University of Algarve research partnership. Target: test report sufficient for municipal approval under "innovative construction system" pathway.

→ Active

Ground screw capacity in Algarve soils

Helical ground screws are proven in Northern European soils (clay, sand, loam). Algarve soils are variable: sandy near coast, rocky inland, clay in valleys. The pull-out capacity of a 68mm screw in sandy Algarve soil has not been tested for this project.

Mitigation: On-site pull-out tests before foundation installation. Increase screw count in low-capacity soils. Alternative: shallow pad foundations in bedrock areas.

→ Active

Fire rating

Target: REI 30 (30 minutes resistance to fire, insulation, and integrity). Raw timber + OSB + rock wool provides natural fire protection through char layer formation. But the specific assembly has not been fire-tested. A gypsum board lining on the interior face would significantly improve rating.

Mitigation: Full-scale panel burn test scheduled. Interior gypsum lining as standard specification. Fire stops at panel joints.

? Open

Long-term creep and settlement

Wood creeps under sustained load. A wall panel under roof dead load + snow load + wind load will deform over decades. The magnitude of this deformation in an RCP panel is unknown because the assembly is new. We assume the OSB skins control long-term deformation, but this is an assumption.

Mitigation: 12-month prototype monitoring includes LVDT deflection sensors on critical joints. Long-term data will inform design adjustments for future builds.

The 37.5m² Shell A module, Algarve, Q3 2026

This is not a demonstration. It is a test. The module will be built, instrumented, monitored for 12 months, and then either validated or modified based on the data. Every result, positive, negative, or inconclusive, will be published. We take our inspiration from WikiHouse, who proved that digitally fabricated, open-source building systems can generate real community momentum through technical transparency.

Structural testing (pre-occupation)

Test Standard / Method Target Status
Panel compression test EN 408 / adapted for composite panel Verify capacity >10x design load Planned Q3 2026
Panel shear test EN 594 / adapted for RCP Verify shear capacity >5x design load Planned Q3 2026
Wall racking test EN 594 full wall panel Lateral deflection Planned Q3 2026
Floor load test EN 1195 / I-joist span table Verify 1.5 kN/m² live load capacity Planned Q3 2026
Ground screw pull-out On-site load test, 25 kN target Verify foundation capacity per screw Planned Q3 2026
OSB bond durability Accelerated ageing + shear test No delamination after 10 cycles Planned Q4 2026

Thermal and environmental monitoring (12 months)

Parameter Sensor type Location Sample rate
Air temperature DS18B20 digital Each room, roof void, exterior Every 5 minutes
Relative humidity SHT30 / SHT40 Each room, wall cavity, roof cavity Every 5 minutes
Wood moisture content Resistance-type pin Pallet core in 6 locations Hourly
Structural deflection LVDT displacement Mid-span floor, wall base, roof ridge Hourly
Solar generation Inverter output (Shelly EM) PV array AC output Every minute
Energy consumption Shelly EM clamp meters Main circuits (lights, sockets, heating) Every minute
Wind speed + direction Anemometer Ridge height Every minute
Indoor air quality CO2 + VOC sensor Living space Every 5 minutes

All data published in real time at data.hestiafoundation.org (coming Q3 2026). Raw CSV downloads available. Open-source dashboard code on GitHub. You can fork the analysis, disagree with our conclusions, and publish your own.

20+ sensors. Real-time. Public.

The monitoring system is designed to answer specific engineering questions, not to produce pretty charts. Every sensor has a purpose. Every data stream feeds back into the design.

Structural health monitoring

  • LVDT displacement sensors at floor mid-span, wall base, and roof ridge. Measures creep, settlement, and seasonal movement. Calibrated against initial zero reading post-construction.
  • Strain gauges on I-joist flanges at maximum moment locations. Verifies that stress levels remain within elastic limits under all load combinations.
  • Bolt torque indicators at panel connection points. Detects loosening due to wood shrinkage or cyclic loading.
  • Accelerometer on roof structure. Characterises wind-induced vibration and dynamic response.

Thermal performance verification

  • Heat flux plates on interior and exterior wall faces. Directly measures U-value of the complete assembly (not just the materials).
  • Temperature gradient probes through wall thickness. 5 points from interior face to exterior face. Shows thermal bridging, if any.
  • Infrared camera survey at 3, 6, and 12 months. Visual identification of thermal anomalies.

Moisture and durability

  • Wood moisture content pins embedded in pallet cores at 6 locations (high and low, north and south faces, roof, floor). Detects moisture accumulation before decay can begin.
  • Relative humidity sensors in wall cavities and roof void. Verifies that the breathable assembly manages humidity correctly.
  • Rain detection sensor on exterior. Correlates weather events with cavity humidity response.
  • Groundwater level (if applicable). Monitors foundation moisture exposure.

Energy and occupancy

  • Solar PV production (Shelly EM). Every minute. Verifies off-grid viability.
  • Battery state of charge (BMS data). Tracks storage performance through seasons.
  • Circuit-level consumption (Shelly EM clamps). Lighting, sockets, water heating. Identifies the actual energy demand of a Hestia home.
  • Indoor CO2 and VOC. Verifies that the vapour-open, naturally ventilated assembly maintains healthy indoor air quality without mechanical ventilation.

Data publication protocol

  • Raw data: uploaded to GitHub daily as CSV files
  • Live dashboard: data.hestiafoundation.org, updated every 5 minutes
  • Monthly reports: analysis and interpretation published as Markdown documents
  • Annual report: comprehensive performance summary after 12 months
  • All data: CC0 (public domain). Use it for your thesis, your paper, your competing project. We do not own the data. The building owns it.

What we tried that did not work

This section will grow. Every failed prototype, every incorrect assumption, every design mistake will be documented here. We do not hide failures. They are the fastest way to improve.

Iteration 0: EPS core (abandoned)

Our first composite panel design used an EPS (expanded polystyrene) core between OSB skins, essentially a conventional SIP. This was abandoned for three reasons:

  • Shear failure: EPS provides negligible shear transfer. The panel behaved as two independent thin plates. Under racking load, the skins buckled independently at less than 30% of predicted capacity.
  • Screw pull-out: Standard wall fixtures (kitchen cabinets, shelving, curtain rails) pulled out of EPS at loads under 100 N. The wall could not be used as a wall.
  • Carbon: EPS is petroleum-based and not recyclable at end of life. It contradicts the low-carbon principle.

The timber-core RCP replaced EPS entirely. Shear capacity increased by approximately 20-50x. Screw pull-out increased by 10-20x. And the material is carbon-storing, not carbon-emitting.

Iteration 1: Pallet-only walls without OSB (abandoned)

We tested a wall built from unskinned pallets stacked and bolted together. The structural performance was inadequate: the pallet joints created discontinuities in the load path, and the wall racked significantly under lateral load. OSB skins are not cosmetic. They are structural. They create the diaphragm that makes the wall rigid.

Iteration 2: Nail-only fastening (revised)

Early prototypes used nails only, no adhesive. Panel delamination occurred under cyclic humidity loading, the OSB shrank and swelled, working the nails loose. The addition of continuous PU adhesive bond solved this. The current specification uses both: adhesive for continuous shear transfer, nails for clamping pressure during cure and mechanical backup.

This space will be updated. The prototype build in Q3 2026 will produce new failures. They will be documented here, with photographs, data, and design changes. If you are evaluating whether to trust this system, watch this space. A project that has no documented failures has either not built anything, or is hiding something.

Thesis opportunities in a field with zero peer-reviewed literature

No academic paper has ever been published on pallet-core composite panels for housing. That means every question is open. Every test is novel. Every student who works on this is writing the first draft of a new field.

Civil engineering

  • Composite panel shear capacity testing
  • Long-term creep behaviour of timber-OSB bonds
  • Diaphragm action in pallet-core wall systems
  • Seismic performance of lightweight timber assemblies
  • Ground screw pull-out in Southern European soils

Architecture

  • Spatial quality of small-footprint homes
  • Natural ventilation design for vapour-open assemblies
  • Daylight optimisation in narrow-plan modules
  • Biophilic design with exposed timber surfaces

Environmental science

  • Embodied carbon comparison: RCP vs SIP vs timber frame vs masonry
  • Life cycle assessment of pallet-sourced building materials
  • Indoor air quality in naturally ventilated, low-VOC assemblies
  • rainwater harvesting system design for off-grid homes

Computer science / Data

  • Open-source sensor dashboard development
  • Machine learning for building performance prediction
  • Parametric CAD models for automated panel generation
  • Blockchain-verified supply chain tracking for pallet sourcing
Contact research team →