One of our customers runs two automations. For the same fleet. The ISO 50001 fleet energy balance should come out of both. One reads the fuel card statement every morning. Plate. Liters. Euros. Timestamp. The other pulls telematics positions every hour. Kilometers. Vehicle. Shift. Both run reliably. Both write their own log. But the two datasets have never met. No join. No shared table. No number that comes out of both.
That gap is exactly where ISO 50001 and ISO 14001 point.
ISO 50001:2018 requires a documented energy review in section 6.3. Three steps. In a fixed order. First: analyse consumption. What is bought. What is consumed. When and where. Second: determine the significant energy uses. For a forwarding company that is almost always diesel. Not the electricity in the office. Third: identify improvement opportunities. With an estimate of future consumption.
Section 6.4 builds on that. It requires energy performance indicators. EnPI, short for Energy Performance Indicator. One indicator per significant energy use. Trackable over time. For a fleet that is liters per 100 kilometers. Or kilowatt-hours per tonne-kilometer. Section 6.6 requires a plan for the data collection itself. The measurements must be accurate. And repeatable. Not estimated once a year.
ISO 14001:2015 requires the same principle. From the environmental side. In section 9.1. Monitoring. Measurement. Analysis. Evaluation of environmental performance. With valid methods. With calibrated equipment where needed. Both standards want the same thing in the end. A defensible number. Not an estimate from the quality manager's memory.
This is not a niche topic. The ISO Survey 2024 counts 38,482 valid ISO 50001 certificates worldwide. And 676,232 valid ISO 14001 certificates. Both figures from the IAF CertSearch database. More and more customers now ask for an energy or CO2 indicator. As part of the tender. Not as a nice-to-have anymore.
Back to the two automations. Over one calendar week, from September 8 to September 14, the fuel card pipeline read 4,347 rows from the statement. 690 of them matched cleanly to a vehicle in the fleet master data. Together roughly 136,000 liters. Pure diesel. The significant energy use under 6.3.
The telematics pipeline ran in the same window. Independently. 77 vehicles in the active fleet. A single run on September 14 produced 59,088 position logs. Condensed into 92 shifts. Both series of numbers are clean data. Both are fully logged. But no fuel card row knows which shift it belongs to. And no shift knows what the tank used that day.
The missing step sounds trivial once you name it. A join by plate and day. Yet it does not exist as a running process at most forwarding companies. It exists once a year. Right before the audit. As a spreadsheet.
The indicator itself is the smaller half of the job. The larger half: figuring out which row counts at all. In the same week, the fuel card pipeline dropped 3,657 of 4,347 rows. Not as errors. As three clean cases.
First case: a fueling with no vehicle in the master data. 1,220 rows. A plate that is not in the fleet master. Pool vehicles. Rental cars. A shared account for a workshop. Every row needs a decision. Does it belong in the fleet energy balance, or not.
Second case: a car instead of a truck. 168 rows. Private vehicles on the same fuel card. Under 6.3 they do not belong to the fleet's significant energy use. They belong in a different balance.
Third case, technical rather than substantive: 1,726 rows were simple duplicates. The fuel card export delivers a rolling window. It overlaps with the previous day. Not a plausibility case. Just a reason why a raw row count alone means nothing.
Add a fourth case. One that only shows up on the telematics side. A trip with no matching fueling. It often happens because a vehicle moved without a driver card inserted. The telematics then logs no shift. Even though the toll system or the fuel card shows a real movement. The reverse case exists too. A fueling with no trip. A canister for a generator. A misbooking. A second vehicle at the same pump. Every one of these four cases needs its own row with a reason. Not a lumped deviation.
Once these four cases are cleanly flagged, the rest is arithmetic. For every vehicle and every month: liters from the fuel card. Kilometers from the telematics. Divided and multiplied by a hundred. That is the EnPI from 6.4. In its most common form. Liters per 100 kilometers.
Liters convert into an energy amount. Diesel delivers 42.8 gigajoules per tonne. Per the official standard values of the emissions reporting regulation. At a density of 0.845 tonnes per 1,000 liters. That comes out at roughly 10.05 kilowatt-hours per liter. The same regulation gives the CO2 value. Roughly 2.68 kilograms of CO2 per liter. Three indicators from two data sources. One join. One conversion. None of this is newly invented. What is new is only that it happens automatically. Month after month. Instead of once a year by hand.
This is explicitly not a shipment-level number. Anyone who needs a CO2 value per transport under ISO 14083 works with the same raw data. But on a different level. Per shipment. With empty runs and well-to-wheel factors. The energy balance in this post stays at the level of vehicle and site. For the management system itself.
For the workshop and the fleet manager, the order changes first. The fleet master data has to be correct before the indicator can be correct. A new pool vehicle. A decommissioned vehicle. A changed fuel card. Every one of these changes has to reach the master data. Or the join produces another unmatched row.
For the environmental or quality manager, the audit day changes the most. Instead of pulling liters from one portal and kilometers from another for a week, they export a table. Indicator per vehicle. Per month. With a reason column for every row that was not clean. For sustainability reporting, whether CSRD or a customer request, it is the same table. Just with a different heading.
Three honest limits.
The fleet master data has to be maintained. No automation tool notices on its own that a plate now belongs to a different vehicle. That stays a human task. Just a smaller one.
Electric and gas vehicles need their own calculation. The liters-and-density conversion applies to diesel. A vehicle with a different drivetrain needs its own EnPI. Not an approximation borrowed from the diesel factor.
And the automation delivers the indicator, not the judgment. Whether a rise in consumption comes from the route, the driver, the vehicle's age, or the load, that is a professional question. For ISO 50001 the indicator is enough as a starting point. The interpretation stays with the fleet manager. Not the algorithm.
A fuel card ingestion with fleet master matching. A telematics pipeline with shifts per vehicle and day. A join by plate and day, with four named exception cases instead of one lumped deviation. A conversion following official standard values for kilowatt-hours and CO2. An export as an EnPI table, for sections 6.3, 6.4 and 6.6, and equally for ISO 14001 section 9.1.
Do your fuel card and your telematics already meet in one table, or only in the environmental manager's head right before the audit? Tell me which vehicle in your fleet would show the biggest gap. Then I will show you, in a process check, what the join looks like for exactly that fleet.
