Energy per unit made, with the base load taken out.
Utility and Resource Intensity measures each metered resource at a site, electricity first, then water, steam, compressed air, fuel and gas, per tonne, litre or kilogram made, half hour by half hour. It works out what the site should have used for what it actually made, and shows the gap. Making more or less stops looking like performance.
- Expected for what was madefrom the previous four weeks354,718 kWh
- Actually usedsite mains, every half hour350,667 kWh
- Gapwithin the baseline, not a saving to claim−1.1%
kWh per unit falls when you make more. That is not efficiency.
A plant draws a base load whether or not the line runs: compressors, chillers, lighting, machines powered and waiting. Divide the month's electricity by the month's output and that base load is spread over however many units were made, so the ratio reports volume as if it were performance. The only way to know what a unit costs is to separate the two loads, and that needs resource and production measured together, at the same resolution.
On the food manufacturing operation, two lines stopped over three days in the same week. Output fell from 706.9 to 254.8 to 50.8 tonnes a day, and energy per tonne rose from 67.8 to 105.9 to 320.1 kWh. The lines that kept running did not get worse. The base load landed on almost no output. A bill that says "kWh went up" cannot tell those apart; the baseline can.
Method: Specific energy consumption · Fixed and variable energy
Site resource against site production, in two dimensions.
Augos meters the resource at the site mains and counts production on every line, expressed in one physical unit the whole site shares: tonnes, kilograms or litres. A pack is not a unit of resource cost, because packs of different sizes do not cost the same to make. Then it splits the result two ways.
- Base load and variable load: A regression of resource on production separates the fixed draw from the amount each unit adds. The base load has a facilities owner; the slope has a production owner. Both are reported, and the ratio is shown alongside them for what it is.
- Operating and non-operating: Paid time from the shift plan decides which half hours are operating. Non-operating consumption is what the site draws when nobody is scheduled; operating base load is what it draws during a shift while the lines stand still, the same idle time the Owner column already attributes. The boundary is editable when production is recorded outside the plan.
- Any metered input: Electricity first, because every site meters it. Then water, steam, compressed air, fuel and gas, and any counted input that goes into the product: carbon dioxide on a beverage line, flavourant on a snack line. Consumed supply is what counts; deliveries by truck are a stock movement. Packaging and raw materials per unit come next on the same method.
- Half hour, day, week, month: The working resolution is the half hour, the interval South African electricity is billed and settled at, which is fine enough to see a shift ramp up, a break, and the tail after the last line stops. Day, week and month views roll up from it.
Four views of one week.
Each view answers a question a plant manager asks on Monday morning. All four come from the same half-hourly data, so they never disagree.
- Expected against actual: The baseline is fitted on the previous weeks and applied forward, so the week is scored against what the site normally needs for that output, not against itself. The gap is the number to investigate: a stoppage, a slow shift, equipment drifting, or a real improvement.
- The half-hour map: Every half hour of the week in a grid, days down, time across, shaded by energy per tonne. Time-of-day dips and bad days show at a glance. Half hours with no production are marked as stopped, never as efficient.
- Speed against efficiency: Energy per tonne grouped by production rate. On the food manufacturing operation it held at 62 to 71 kWh across normal running rates and averaged 174 in the slowest tenth of half hours. Running slower to save power costs more per tonne, not less.
- The best half hour: The day's most efficient half hour sets that day's target, and the report counts how many half hours ran within 15% of it. On one production day it was 31 of 48. The worst hour was a midday dip where output fell by two thirds and consumption barely moved.
Augos platform, one anonymised food manufacturing operation, 2026. Half-hourly production reconstructed from each line's job tonnage and recorded stops, matched to the site mains. One meter serves the whole site, so these are site figures, not per-line figures, until lines are sub-metered.
Every stop has an energy price.
When a line stops, the base load keeps running: compressors, extraction, lighting, the rest of the site. OEE+ already records every planned stop, unplanned stop and interruption, with an owner. Resource intensity puts a figure on what those minutes cost in kWh per tonne, from the same counts, so a stop that looked cheap on the production report shows up on the energy one too.
A second plant, a different lesson.
Output per production day was the same all month. Electricity per production day fell by 36% in the last ten days, so the simple ratio fell from 4.61 to 2.94 kWh per thousand bags and would have been read as an efficiency gain. The fixed-plus-variable view puts the change where it belongs: in the base load, on a single date. What was switched off is the plant's question to answer; the data says only that production did not change and the standing load did. On the half-hourly meter the operating day reads as a ramp from 06:00, a plateau of 33 to 38 kWh per half hour until 13:30, an afternoon at 24 to 26, and a return to 2 kWh per half hour by 17:00.
Augos platform, one anonymised flexible-packaging plant making one bag size, one 30-day month in 2026. Daily figures; the module's working resolution is the half hour. No outcome is claimed; the plant's cause is unknown to Augos.
See what a unit really costs on your site. We put your site's meter against your production and show you the base load you pay before the first unit is made.
Your Scope 1 and 2 are your choice. Your customer's Scope 3 is becoming a trading requirement.
A manufacturer's measured energy per unit is its own Scope 1 and 2 intensity, and it is the supplier-specific input a customer needs for Scope 3. Under a spend-based or average-data method nothing the plant does to its energy use changes the customer's number; under a supplier-specific method, a plant that cuts its measured kWh per unit cuts its customer's footprint. Buyers are starting to ask for that data. The plant that can hand over a measured per-unit figure, with its boundary stated, is ready when they do.
Method and context: Scope 1, 2 and 3 for a factory. Augos provides measured activity data; emission factors, allocation and assurance belong to the plant's carbon accounting.
The same sites, the same meters.
- The resource side: Electricity, water, gas, fuel and compressed air metered at the site. Electricity meters record every 60 seconds; every resource is rolled up to the half hour. Augos Energy →
- The production side: Every unit counted at the pulse sensor and the Augos Operator Terminal, every stop owned. The idle time the operating base load pays for is the same time OEE+ attributes. OEE+ and the Owner column →
- The method, written down: What specific energy consumption is and why the ratio misleads, how base load and slope are separated, and where a per-unit figure sits in the three scopes. The Factory reference pages →
- Yours to take out: Every figure exports as CSV, Excel or the signed AI Bridge payload, scoped to the measurement points you are cleared for. AI and connectors →
What it costs, and what you get.
Utility and Resource Intensity is included in the standard fee: R4,950 per line per month, plus R19,500 once per line for deployment, ex VAT. Resources measured with Augos counters and sensors are charged at their standard prices.
R19,500 once per line for deployment, ex VAT.
Several lines? The price per line falls. We agree it with you in conversation.
Book a demoThe first 90 days cannot be cancelled. After that, one calendar month's notice, and the hardware comes back to Augos.
The monthly fee includes
- Every capability, except the metrics that need Augos Assist
- The Augos Operator Terminal and pulse sensor, owned by Augos
- A SIM supplied by Augos, so there is no network or IT project
- A free terminal replacement every 36 months, and a one-year warranty
The once-off fee covers
- Set-up and remote testing
- Two days of on-site operator training
- Four weekly one-hour data reviews
- Leadership training on the platform
Travel and subsistence for on-site training are not included.
What you provide
A power point at the line. Your own people mount and connect the terminal, with our guidance.
Optional extras, per month, ex VAT
- Augos Assist: R250 per technician, or R1,250 for the whole site
- Public Display: R300 per display, or R1,500 for the whole site
Common questions.
What is Utility and Resource Intensity?
An Augos Factory capability that expresses each metered resource at a site, electricity first, then water, steam, compressed air, fuel and gas, per unit of the site's production, half hour by half hour, against a fixed-plus-variable baseline. It answers how much of a resource each unit really costs, and how much the plant pays just to be switched on.
Why not just divide monthly kWh by monthly output?
Because a plant's base load is spread over however many units were made, so the ratio falls when volume rises and rises when it falls, regardless of efficiency. On one anonymised packaging plant the ratio fell 36% in a month at constant output because less equipment was running. Splitting base load from variable load shows that.
Do we need to meter every line?
No. Augos starts at site level: the site's mains meter against the site's total production, expressed in one physical unit such as tonnes, kilograms or litres. A count of packs or bags only works where every unit is the same size. Shared services such as compressors and chillers serve every line, so site level is where the arithmetic is honest. Sub-metering refines the picture later.
Does running the line slower save energy?
Usually not. On one anonymised food manufacturing operation, energy per tonne held at 62 to 71 kWh across normal running rates and rose to an average of 174 kWh per tonne in the slowest tenth of half hours, nearly three times as much. The base load does not slow down with the line. The saving is in avoiding long stretches of near-idle running.
Where does production come from?
From the production counters or the Augos Operator Terminal on each line, the same counts that feed OEE+. Production is expressed in one physical unit, mass or volume, so that lines making different products can be added honestly.
Is this carbon reporting?
It is the measured activity data a footprint is built on: resource per unit at half-hour resolution with its boundary stated. Emission factors, allocation and assurance are applied by the plant or its advisers. Augos does not describe it as Scope 3 reporting or as compliance with any disclosure regime.
See what a unit really costs.
A 30-minute call. We put your site's meter against your site's production and show you the base load you pay before the first unit is made.
- Live within 48 hours of your kit arriving
- One sensor, one terminal, no PLC and nothing to integrate
- R4,950 per line per month, ex VAT. 90 days, then a month's notice