From the main meter to consumption per machine.
The bill tells you what you spent, never where it went. Meddle attributes kWh to zones, departments and individual machines, builds a baseline that survives a change in product mix, and measures the saving where it actually shows: on the invoice.
- Per machinenot per site: kWh attributed where they are spent
- Baselinenormalised on volumes, so the comparison holds
- On the billsaving verified against supplier data, not estimated
Why energy stays a cost you absorb
A single measuring point
The main meter returns a total, and a total cannot be optimised: it does not say which line, which shift or which machine produced it, so every intervention starts from a guess.
The baseline breaks on the first mix change
Comparing March with February means nothing if the volumes differ. Without normalising on units produced, a real saving disappears and a real regression goes unnoticed.
The saving stays on paper
The work gets done, the report estimates a percentage, and nobody goes back to verify it against the data. When it is time to fund the next project, there is no evidence to show.
Three steps, from measurement to verified saving
Measure where it matters
Meters, drives, PLCs and the BMS come in as sources. Where no measurement exists, an instrument is added — only on the loads that genuinely weigh.
Attribute and normalise
Every kWh gets plant context: area, line, machine, shift, machine state. The baseline is built per unit produced, not per calendar month.
Verify on the invoice
Post-intervention consumption is compared with the baseline at equal volumes, and the delta is reconciled against supplier data. The resulting number is defensible.
Plain-language questions, answers from your own plants
- What did compressed air cost us last month, by line?Compressor consumption allocated across the lines it serves, converted into euros at the period's supply price. Circuit leaks read directly from consumption while production is stopped.
- Which machines consume most at equal output?kWh per unit, machine by machine — the comparison that matters. High draw on high volume is normal; the same draw on low volume is a problem.
- Did April's HVAC work actually save anything?The post-intervention period compared with the normalised baseline, with degree-days as the weather correction. The answer comes back in kWh and in euros.
- How much are we drawing right now, and how much is outside production?Real-time draw with line state beside it, so idle consumption — usually the easiest item to attack — is visible immediately.
What an audit needs, not just a dashboard
- Intact history. Data stays at the granularity it was collected at and is exportable at any time. A baseline is only defensible if its reference period can still be inspected.
- ISO 50001 compatible. Baseline, EnPI and periodic reporting are already the shape the data comes out in, with no second system to keep aligned by hand.
- No lock-in. Standard protocols in, your data out. If you change platform, the historical series stays yours and stays readable.
Let's start with the loads that weigh.
A guided pilot on compressors, HVAC and your two most energy-hungry lines: within weeks you have the real consumption breakdown and a baseline to measure the first intervention against.
Frequently asked questions
Do we need new meters on every machine?
Rarely. Most of the draw is already measured by existing drives, PLCs and switchboards, and is reachable in software. An instrument is added only where measurement is genuinely absent and the load is big enough to justify it.
How soon do we see the first consumption breakdown?
Sources that are already measured come in within days, not months. The breakdown by area is readable as soon as there is enough history to separate the shifts; a solid normalised baseline needs a few weeks of production.
Does it work with twenty-year-old machines?
Yes. The constraint is not the age of the machine but the presence of a measuring point: a PLC exposing Modbus, or a drive with its own interface, is enough. Where there is nothing, measurement moves upstream to the switchboard.
Is the data usable for energy-efficiency incentive schemes?
The kind of data is the kind these schemes ask for — consumption before and after, with a documented baseline and verifiable history. The filing itself remains a matter for your consultant and certifier: the platform supplies the measurement, not the attestation.
How do you tell a real saving from a drop in production?
By normalising. The comparison is on consumption per unit produced, not on the monthly total, and a degree-day correction separates the effect of the season from the effect of the intervention.
Let's find out where your energy actually goes.
In a guided demo we start from your own loads and your own consumption, not from an example.
Book a demo