A passport is not a document but a data set. Data whose collection starts when the obligation arrives can never be completed retroactively.
In most organisations, the digital product passport discussion starts with "when will it become mandatory?". A more useful question is: when it does, which data can we produce retroactively?
The answer is usually "very little". Which batch a product came from, which raw materials it was made with and which facility it passed through cannot be reconstructed later if they were not recorded at the time. A document can be produced; the data cannot.
The four layers a passport carries
- Identity: the product's unique number and the mark that makes it scannable.
- Composition: raw materials, material ratios, share of recycled content.
- Origin: production facility, batch, date and, where relevant, the supplier chain.
- Life cycle: maintenance, repair, reuse and disposal instructions.
Without the first layer, the other three have nothing to attach to. The passport is tied to the UNIT, not to the product model: two batches of the same model can carry different raw material ratios, and the whole point of the passport is to be able to show that difference.
The passport is tied to the unit, not the model; a record kept at model level does not answer the question being asked.
Three things you can do today
First, start unit identity. Even if your label carries only a barcode today, giving each unit a unique number is the most expensive layer to add later.
Second, link batch information to that identity. It is already in your production records; what is missing is the link between that record and the number on the product.
Third, match supplier documents to batches. Analysis reports, certificates of origin and declarations of conformity sit in folders today; the passport wants them linked in a machine-readable way.
The return on starting early
An organisation that starts collecting data early faces not a software project but a format conversion when the obligation arrives. Converting the data you have into the required schema takes weeks; producing data you do not have is impossible.
Where the data comes from
Most of the information the digital product passport asks for already exists in the plant: raw material sources, production dates, bills of materials, certificates. The problem is not that the information is missing but that it is scattered. The same information sits separately in the quality file, the supplier contract and the label approval, and one day the three drift apart.
The first step in preparing is not collecting new data but linking what exists to a single source. Unlinked data is compiled by hand when the passport becomes mandatory, and every hand-compiled file is out of date within the first month.
At which product level is it kept?
Whether the passport is tied to the product model, the batch or the individual unit is a decision to make early. A passport tied to the model is cheap but cannot describe batch differences; one tied to the unit tells the most but requires every unit to have its own identity.
- Model level: enough for information that does not change across the same product.
- Batch level: needed when raw materials and production conditions change.
- Unit level: needed if recalls and ownership tracking are required.
- The level can be raised later; it cannot be lowered.
When is the passport updated?
A passport is not a document published once and left alone. Some information changes over the product's life: repair, resale, recycling. If it is not decided up front who can update it and how old versions are kept, the passport drifts away from the product's actual state over time.
The data you will need to collect when the passport becomes mandatory is the data you are not collecting today.
