DPP Madera
Proof of concept · engineered timber buildings

A digital passport for every timber part, from factory to finished building

Today, what happens to a timber panel is scattered across delivery notes, spreadsheets, photos and emails. DPP Madera gives each part one identity, scanned from a QR label, and links four simple records to it. Anyone with the right access can then ask a question about a part, a floor or the whole building and get one answer.

The idea in one minute

1. MadeThe factory cuts a CLT panel and prints a QR label with its identity.
2. MovedEvery hand-over (lorry, yard, crane) is recorded against the same identity.
3. CheckedSite staff scan the label and add readings, such as moisture.
4. InstalledThe part is linked to its place in the building model.
5. SeenA live picture shows progress and raises a flag when something does not add up.

No single company has to own all the data. Each record stays with whoever is responsible for it, and the records point to each other instead of copying.

The four records, in plain words

DPPDigital product passport

The part's ID card. What it is, which batch of timber it came from, who made it and who has had it since.

Example: floor panel P03, cut from timber batch L27-01546 on 4 January, dispatched on 6 January.

EPMSite observations

Measurements and checks made on the part, by a person or a sensor. About parts only, never about workers.

Example: two moisture readings on P03 that disagree (11.8% and 18.5%).

DBLDigital building logbook

The building's own record. Where each part sits in the building, as actually built, linked to the design model (IFC).

Example: P03 sits on level 1, zone L1-E, and matches one element of the model.

DTwAssembly twin

A live picture of progress. It holds no data of its own: it reads the other three records and shows the state of every part.

Example: 18 parts inspected, 1 installed, 1 delivered, 6 flagged for a closer look.

Find your way in

The same building is shown five ways. Pick the one that matches who you are.

What the proof of concept shows

11 / 11tests pass, from identity resolution to a one-command rerun
5 / 5deliberately planted faults caught (missing hand-over, conflicting readings, model clash, wrong order, late record)
1request answers a question that would take 4 to 40 manual look-ups (estimate)
0facts stored twice: each record keeps its own data and points to the others

Readiness: the project enters at TRL 2. This proof of concept is strong evidence for that level and the first step towards TRL 3, which the project plans to reach at month 6 using real data from partner companies.