The questions people actually ask us.

Both, and they are inseparable. The unit fetches the data; the subscription covers what is done with it — reading, a written report, adjustments on site. A unit sold on its own would send back data nobody reads after three months.

No. DConnect reads what is already there. On one measured site, six identical dryers showed up to 25 % of variation between them — with no machine at fault. The benchmark was already in the shop.

That is common. Two answers: fit our own metering points, or read visually what is displayed — a dial, a lamp, a product passing — with the Vision module. A camera measuring point costs €190-290 in hardware.

No, and that is a design decision. The link is read-only: DConnect never writes to your process. Adjustments are made on site, with your agreement and under your lock-out procedures.

Because it depends on how many energy uses need instrumenting, how accessible the measuring points are and how often we come — three things only known after the site visit. The quote is given line by line, with no opaque lump sum.

Acquisition and history stay on site, in the unit in your plant room — a network outage leaves no gap. Our remote access runs through a dedicated encrypted tunnel, with no port opened on your industrial network. Hosted in France.

You leave with the complete history, in a standard format readable without our tool. Not a PDF summary: the raw data. It is written into the contract, and remote access can be revoked by you at any time.

By opening it. For any figure on screen, the formula, the source quantities, the period and the assumptions can all be inspected. If you cannot redo the calculation yourself, the indicator is worthless.

Rarely. First the sampling interval: a cycle lasts a few tens of minutes, and measured every quarter of an hour it disappears. Then the link to output: raw consumption rises when output rises, and without tonnage alongside it tells you nothing. That said, instrumentation already in place is reused wherever it can be qualified.

The first written report comes after about a month. But we do not conclude yet: a month proves the measurement is sound, not that a trend exists. The first stable gaps appear around the second or third month, and telling drift from a seasonal effect takes a full year.

No, and we stand by that. Quoting a percentage before measuring a site amounts to inventing it. What we publish are findings from real installations, with their scope and their assumptions — never a promise you can transpose. See the readings →

Little once it is running, more at the start. What we mainly need is a named contact on site with the authority to decide. That is the real success factor: a measurement project rarely fails on the technology, it fails when nobody has the mandate to enforce a correction.

Yes, with a consolidated console for comparing sites. Each site keeps its own measurement plan: indicators are only comparable if the scopes are, and we refuse comparisons that do not hold.

Yes. It falls under the French standardised scheme IND-UT-134 and meets its technical requirements: a sampling interval far below the maximum allowed, a rolling year of history in direct access, dedicated software, automatic drift alerts, and installation by a professional.

We produce the technical documents — which the work generates anyway. The administrative filing is handled by a partner agent. That split is deliberate: it keeps us independent on the technical substance.

We do not publish figures. The volume depends on the rating of the monitored equipment, the operating regime and the length of the contract — and the final amount depends on the certificate market, which moves. An estimate is provided with the proposal, with the calculation shown.