This project uses embedded sensors and a microcontroller to monitor the conditions inside some Hemp Crete walls.

The Curing Prediction Engine runs server-side using a multi-factor drying rate model. Historical moisture readings are normalised against concurrent temperature and humidity data to derive a baseline drying rate for each specimen. This rate is blended with recent observations and adjusted for wall aspect, sensor position, and a 14-day weather forecast to produce a forward projection. Confidence is assigned based on data density and projection plausibility.

Each hempcrete wall develops a unique thermal fingerprint as it cures — the specific way it absorbs, delays, and releases heat. Three metrics describe this signature:

Lag — hours between the outdoor peak and indoor peak. Longer = the wall holds heat longer before releasing it inside.

Damping — how much the wall reduces the daily temperature range. Higher is better for comfort.

Decrement — interior swing ÷ exterior swing (ISO 13786). Closer to 0 is better. Dry hempcrete typically scores 0.3–0.5. Each wall's signature develops as it cures.