For years, a fleet's Scope 1 emissions number was whatever a spreadsheet said it was: mileage multiplied by an average emissions factor for the vehicle class. Nobody asked how it was calculated, because nobody was checking. CSRD changes that.
Estimates were never the problem, until now
A GPS-based estimate infers fuel burn and emissions from distance travelled and a generic vehicle profile. It's a reasonable proxy when the number is informational. It's a weak foundation when the number has to carry limited assurance from an external auditor, which is what CSRD requires of the sustainability statement.
An estimate can't show its working. It can't reconcile against a specific vehicle's actual fuel or energy consumption. And it treats a diesel van driven hard in stop-start traffic the same as one cruising a motorway, when the two burn meaningfully different amounts of fuel per mile.
What manufacturer-grade data changes
An OEM native API reports what the vehicle itself measured: actual fuel or energy consumed, not a distance-based inference. That figure can be traced back to a specific vehicle, on a specific day, and reconciled against fuel card or charging data for the same vehicle. It's the difference between an emissions figure you calculated and one you can defend.
An auditor doesn't ask whether your CO₂e number is directionally right. They ask how you calculated it, and whether you can show the underlying data. Only one of those two methods survives that question.
The practical shift
This doesn't mean re-platforming the whole fleet. It means the emissions line in your sustainability statement should be built on actual consumption data wherever a vehicle can provide it, with GPS-based estimation as the fallback for the vehicles that can't, not the default for the whole fleet. As CSRD assurance requirements tighten each reporting cycle, that distinction is exactly what gets tested.