Product data is cheapest on the day the product is made

A customer’s compliance team sends a questionnaire. For a product you shipped last spring, they want the material composition, where the main components came from, which batch of a particular raw material went into it, and a few process parameters from the day it was made.

Every one of those facts existed on the day the product was made. It was on a delivery note, in a supplier certificate, on a machine display, in an operator’s head. Assembling them now takes a week of somebody’s time, and the result is a careful reconstruction rather than a record.

Where European product rules are heading

The direction is towards structured, machine-readable data attached to products, rather than documents filed about them. The Ecodesign for Sustainable Products Regulation (ESPR) introduces a digital product passport, with the detailed requirements set product group by product group in delegated acts — including whether the passport relates to a model, a batch or an individual item. The EU Batteries Regulation sets out a battery passport for certain categories of battery.

The detail will arrive group by group over years, and much of it has not been written yet. A plant does not need to know exactly which fields its product group will require to see the shape of the problem: at some point, somebody will ask for data about what was actually made, at a finer grain than most plants record today.

Why reconstruction is the expensive way

The facts are scattered. Materials in the ERP, batches on paper, process data in the machine, inspection results in a quality register. Joining them after the fact is manual work, every time.

The links were never made. If the connection between a raw material batch and the units made from it was not recorded at the moment of consumption, no amount of searching later can recreate it with confidence.

The people have moved on. The operator who knows which reel was loaded that morning is on a different shift, or in a different job. Their memory is not a record and should not be treated as one.

It does not scale. Answering one questionnaire by hand is a nuisance. Answering for every unit shipped, as a standing obligation, is a job description.

Capture on the day

The cheaper approach is unglamorous. Record product-level data in a structured form at the moment it exists, as a by-product of running production rather than as a separate exercise for compliance.

Consumption at batch level. Which material batch went into which order, at which operation, recorded when it is consumed.

Process data from where it already is. Parameters and counts read from the equipment that produces them, instead of copied out by hand.

Inspection results on the unit or the batch. Attached to what was inspected, not filed by date.

One object per order. So that a question about a product is a query against one record rather than a search across four systems.

Most of this is the same data a plant needs for traceability and honest costing anyway. That is the point: the regulatory use is one more question asked of a record that is worth keeping for its own sake.

What this is not

It is not compliance. Capturing data well certifies nothing, and any product that says otherwise is promising something no software can give. What good capture does is turn the eventual obligation into a reporting exercise rather than an archaeological one.

A simple check

Take one product shipped six months ago and try to fill in a plausible passport for it: materials, origins, batches, a handful of process parameters. Note what you had to look up, where, and how long it took.

That exercise costs an afternoon now. Doing it for every product later costs considerably more.