Task · homologation
Vehicle homologation data
Approval is granted against evidence. The engineering question is whether your measurements, datasets and decisions can still be reconstructed months later.
In short
Vehicle homologation is the process of demonstrating that a vehicle meets the regulatory requirements of a market before it can be sold, including emissions and on-board diagnostic requirements measured over defined cycles such as WLTP. From a data perspective it requires three things: measurements analysed consistently, calibration datasets whose history can be reconstructed, and a defensible link between the two. VehicleLab supports that locally: MDX analyses the cycle recordings, CIE Pro documents how a calibration optimum was reached, and HELIX freezes a traceable release state that ties a dataset to the runs that justified it.
01
What the data has to prove
An approval file is a claim about behaviour, supported by measurement. Months later someone will ask why a value is what it is, which measurement supported it, and whether the software in the vehicle is the software that was tested. Those questions are answerable only if the trail was kept while the work happened.
The practical requirements are unglamorous: recordings stored in a standard container with their metadata intact, analyses that can be re-run to the same result, datasets whose changes are attributable, and a release state that is frozen rather than described.
- Cycle analysisWLTP drive-cycle analysis over the recording, with phases and operating points identified consistently.
- Trace executionHow closely the run followed the prescribed speed trace, held separately from what the calibration did.
- Diagnostic evidenceMonitor behaviour and in-use ratios verified against real driving, not asserted.
- Dataset provenanceSource A2L, base dataset, approved changes and validation status held together.
- Frozen releaseA state you can return to, with the traceability that makes a later question answerable.
02
Consistency is the whole game
Two engineers analysing the same cycle should reach the same numbers. That means the analysis lives in a script or a saved configuration rather than in a person's method, reads channels at their own rasters, and marks derived quantities as derived. MDX exposes its analysis through a Python SDK and a local REST API precisely so a homologation analysis is reproducible rather than remembered.
On the calibration side, CIE Pro records how an optimum was reached — the designed points, the fitted surface, the constraints applied — so the dataset is not just a set of numbers but a decision with reasoning attached.
Consistency also means being explicit about which measurement chain a number came from. A bag-based carbon-balance result, a controller-integrated fuel rate and a cumulative trip counter do not agree in absolute terms, and a comparison that crosses between them is not a comparison. State the chain, stay inside it, and correct for anything that moved energy around rather than removing it — an electrical charge balance that was not neutral, most commonly.
03
Working where the data cannot travel
Homologation work involves unreleased vehicles and confidential calibration IP. The desktop instruments run fully offline with no licence server, and the agent-based tools deploy on-premise, so an approval programme can run without prototype data leaving the building.
We are honest about scope: VehicleLab provides the measurement analysis, calibration modelling and release traceability. The testing itself, and the regulatory submission, remain with your test house and homologation team.
The instruments
What does what.
- OSCILLOSCOPE LAB & MCPMDXWLTP cycle analysis, residence and event detection, reproducible through the Python SDK and local REST API.→
- DIGITAL TWIN CALIBRATIONCIE ProDocuments how a calibration optimum was reached — designed points, fitted surface, constraints.→
- RELEASE GOVERNANCE GATEWAYHELIXChange packages, validation gates and a frozen release state tying datasets to the runs behind them.→
- NEURAL DIAGNOSTIC GRAPHDiagAIWLTP and diagnostic agents that verify monitor behaviour in the recorded driving.→
Standards & references
Check the source.
These primary sources define the formats and regulatory context discussed above. They do not certify or endorse VehicleLab products.
Questions
Frequently asked.
What is the vehicle homologation process?
It is the process of demonstrating that a vehicle meets the regulatory requirements of the market it is sold in — emissions, on-board diagnostics and other requirements measured over defined cycles such as WLTP — so that a type approval can be granted.
What data is needed for a homologation programme?
Cycle recordings in a standard container with their metadata intact, analyses that can be reproduced to the same result, calibration datasets whose changes are attributable to a source and a reviewer, and a frozen release state that matches the software tested.
Does VehicleLab perform homologation testing?
No. We provide the data side: measurement analysis, calibration modelling and release traceability. The testing and the regulatory submission stay with your test house and homologation team.
Can homologation data stay on-premise?
Yes. The desktop instruments run fully offline with no licence server, and HELIX and DiagAI can be deployed on-premise, which is usually a requirement for unreleased vehicles and confidential calibration data.
Which regulatory programmes does this data work support?
Any programme whose evidence is measurement: WLTP cycle work, real-driving emissions logging, and on-board diagnostic requirements including the Indian AIS-175 and BS6 framework. VehicleLab analyses the data; the testing and the submission stay with your test house and homologation team.
Why does the measurement chain have to be stated?
Because the available chains disagree in absolute terms. Carbon balance from the bags is the regulatory reference; a controller-integrated fuel rate and a trip counter are useful but offset from it. A comparison is only valid inside one chain, and saying which one is part of the evidence.
Next step
Open one of your own files with us.
Send a measurement, an A2L or an EHANDBOOK container. We will open it in the instruments on a call, with your engineers in the room.