White paper: Root of Trust in the software-defined vehicle

Insights · 2026-06-28

What is variant coding? ECU baselines and configuration management

In brief

  1. 01

    An ECU baseline is the approved record of which software, calibration and coded configuration belong in each ECU and variant.

  2. 02

    Variant coding adapts a generic ECU to a specific vehicle, market and option set, written with UDS 0x2E and read with 0x22.

  3. 03

    A single platform spans thousands of variants under continuous change, making configuration management the real engineering challenge.

  4. 04

    ODX (ISO 22901) formally describes the coding data and DIDs so production and aftermarket tools share one consistent definition.

  5. 05

    Drift between the as-designed baseline and the as-built or as-maintained vehicle is the central risk — prevent the gap rather than chase it.

A modern vehicle platform carries thousands of software and configuration variants, each of which must be exactly right in every ECU that leaves the line. Holding one authoritative record of what belongs where is the difference between a controlled programme and a slow drift into the unknown.

What an ECU baseline actually is

An ECU baseline is the authoritative record of what belongs inside a given control unit for a given vehicle variant: which software version, which calibration set, and which configuration values are correct and approved. It answers a deceptively simple question — for this VIN, on this platform, in this market, what should this ECU contain, and how would you prove it does?

The baseline is not the binary in the flash. It is the engineering statement of intent against which the flashed reality is checked. A clean baseline lets you reason about an entire fleet; a missing or ambiguous one means every diagnostic, recall and field repair starts from guesswork.

Variant coding and parameterisation

Manufacturers do not build a unique ECU for every option pack. They build a generic unit and then adapt it — to the vehicle, the market and the chosen options — through coded values written into the ECU. This is variant coding (and the closely related parameterisation of calibration data). A left-hand-drive estate for one market and a right-hand-drive saloon for another may run identical software, distinguished only by their coded configuration.

Illustration: one ECU on a plinth in a dark hall, linked by thin orange lines to three identical unmarked cars
Illustration

In UDS (ISO 14229) terms, those values are addressed by Data Identifiers. Configuration is written with service 0x2E WriteDataByIdentifier and read back with 0x22 ReadDataByIdentifier. The mechanics are straightforward; the hard part is knowing, with certainty, which coded value is correct for which variant — and that knowledge has to live somewhere trustworthy.

Variant coding in practice: one write, one read-back

On the wire, coding one value is a short exchange. Writing is usually only allowed in an extended session and after security access, so the tester first opens the session and unlocks the ECU (with 0x27 SecurityAccess, or 0x29 Authentication on newer ECUs):

  1. Request: 10 03 DiagnosticSessionControl: extended session

  2. Response: 50 03 ... Positive response, with the session's timing parameters

  3. Security access: 0x27 or 0x29

  4. Request: 2E 01 0A 03 05 WriteDataByIdentifier: DID 0x010A, two coded bytes

  5. Response: 6E 01 0A Positive response (0x2E + 0x40)

  6. Request: 22 01 0A ReadDataByIdentifier: DID 0x010A

  7. Response: 62 01 0A 03 05 The coded value, read back

One write, one read-back · bytes in hexadecimal

DID 0x010A is an illustrative manufacturer-specific identifier. What its two bytes mean, say a market code in the first and a bit field of options in the second, is defined in the ODX description, not in UDS. A positive response always carries the service ID plus 0x40. If the ECU refuses, the negative response 7F 2E and a code say why: 0x13 wrong length, 0x22 conditions not correct, 0x31 request out of range (a DID the ECU does not support or will not write in the active session, or a value outside the allowed range), 0x33 security access denied, or 0x72 a failed write to non-volatile memory. Many ECUs apply new coding only after a reset (0x11).

The read-back is the step that matters for the baseline. The value actually in the ECU is compared with the value the baseline says belongs there, and the result is recorded against the vehicle.

The configuration-management challenge

Now scale it. A platform may span thousands of variants — drivetrains, markets, trim levels, regulatory regions, optional features — multiplied across dozens of networked ECUs, and the whole matrix is under continuous change as suppliers revise software and new options appear. Each combination implies a specific, defensible set of coded values.

Without disciplined configuration management this becomes unmanageable. Coding tables diverge between teams, a value approved in engineering never reaches the production tool, and two ECUs that should be identical end up coded differently for no recorded reason. The cost is not abstract: it surfaces as no-fault-found returns, failed end-of-line tests and vehicles that behave subtly wrong in ways nobody can trace.

Where ODX fits in

The coding data and the DIDs it touches are not meant to be tribal knowledge. ODX (ISO 22901-1) provides the formal, vendor-neutral description: the data identifiers, their structure and the ECU configuration records used for variant coding are captured in a single machine-readable model. That same description is meant to flow from development through production and into the aftermarket, so every team works from the same definition rather than re-deriving it.

Because the description is standardised, the production end-of-line system, the workshop tester and the after-sales tool can all interpret the same coded values consistently. ODX is what turns a baseline from a document into something a tool chain can enforce. It pairs naturally with the UDS services that carry out the reads and writes, and with the broader question of how the diagnostic channel is secured.

How programmes keep that description in step with the ECU, authored first, validated automatically and released with the software it describes, is set out in our white paper ODX: one diagnostic truth for the vehicle’s whole life.

Drift is the real risk

The core failure mode is drift: a growing gap between the as-designed baseline and the as-built or as-maintained state in the field. An undocumented production tweak, a workshop that recodes an ECU off-script, a replacement unit flashed with the nearest-looking configuration — each opens a gap between what the records say a vehicle is and what it actually is.

Once the records and reality disagree, every downstream activity degrades. Diagnostics chase phantom faults, recalls cannot be scoped accurately, and cyber-security assurance weakens because you can no longer state with confidence what software is running where. The remedy is not heroic field investigation; it is preventing the gap from opening, by managing baseline and coding from one authoritative source.

This is the problem Diadrom’s DiagStudio is built to address — authoring and holding that single source of truth for diagnostics and configuration, so the same approved definition drives end-of-line and aftermarket alike.

Working on this in a programme of your own? Talk to the team behind this article.

Standards cited here last checked against the issuing bodies on .

Ask us

Have a question about this?

Two fields, straight to the engineers. We reply within one business day.

Subscribe to Diadrom Insights

Brief and technical, once a month.

Diagnostics and secure ECU programming, and the standards that decide what a programme has to deliver. Go deeper with our white paper ODX: one diagnostic truth for the vehicle's whole life.