Book a demo

VehicleLab — Field notesIn pilot

The work, not the logos.

Worked examples from the bench, and what has actually been validated.

We are early, and in pilot with Tier-1 powertrain teams. Rather than dress that up as customer case studies, here is precisely what the instruments do in the situations engineers bring us — and the standards and containers they have been tested against.

Worked examples · not customer results

Validated againststandards.txt
  1. V1Golden containersXHandbook is validated against real Bosch and ETAS EHANDBOOK containers, including the HVR100 and FlexECU demo sets.
  2. V2Open standardsASAM MDF v3 and v4, A2L (ASAP2), Intel HEX and Motorola S-record, DCM, CDFX and PACO — read and written as the standards define them.
  3. V3ProtocolsCAN and LIN bus data, XCP/CCP measurement, UDS (ISO 14229), OBD-II and J1939 in the diagnostic agents.
  4. V4Runs where the data isThe desktop instruments work fully offline; DiagAI and HELIX deploy on-premise when prototype data cannot leave the building.
Facts we can show you on a call
Stage
In pilot with Tier-1 powertrain teams
Worked examples
Measurement, calibration and documentation
Data
Your files, on your machines
Proof
Open your own file on the call

01 — Measurement · cold start

Two cold starts, nine degrees apart.

A start that misbehaves in the cold is a comparison problem: the same eleven signals, two recordings, and the question of where they begin to differ.

  1. 01

    MDX

    Open both runs

    Two MDF4 files on one time base, every channel at its own raster, nothing resampled behind your back.

  2. 02

    MDX

    Line up the cranks

    The runs start at different points in their files. One drag re-times run 2 against run 1, so every later read is at the same point in the start.

  3. 03

    MDX

    Read the divergence

    Cursors report hold-last-sample values and the delta column; catalyst, exhaust, coolant and oil separate where enrichment and retard differ.

ResultYou end with the moment and the magnitude, not an impression of the curve — and the same view opens again next week.

02 — Calibration · dyno time

Fewer test points to the same optimum.

Sweeping a full grid on the dyno spends hours on regions the engine will never run. Design the experiment instead, then let the model say where the next point belongs.

  1. 01

    CIE Pro

    Design the run

    A Design of Experiments over spark and cam instead of a brute-force sweep, bounded by the limits the engine must respect.

  2. 02

    CIE Pro

    Fit and constrain

    A response surface over the measured points, with the knock limit carried as a constraint rather than a line drawn afterwards.

  3. 03

    CIE Pro → Hexx

    Export the map

    The optimum leaves as A2L, CDF or HEX, and lands in the calibration image with its checksums recomputed.

ResultThe value of the method is where the next measurement goes — each point chosen because the model says it is worth measuring.

03 — Documentation · one label

From a label in the docs to the byte in the ECU.

A function description is only useful if you can follow a label from the page to the object it configures — address, layout, conversion, units.

  1. 01

    XHandbook

    Open the container

    An .ehb, .ehbcrypt or .ehbp opens locally: documentation, function models, metadata and search index, no licence server.

  2. 02

    XHandbook

    Follow the label

    From the topic that documents it, through the model block that uses it, down to the A2L object itself.

  3. 03

    Hexx → HELIX

    Change it, traceably

    The map is edited in engineering units against the A2L, and the change travels as a reviewable package.

ResultThe documentation and the calibration stop being two separate worlds held together by a spreadsheet.

Why it reads like this

We would rather show the work than claim a number.

Published case studies need a customer willing to be named and figures both sides can defend. We are not there yet, and inventing them would be the fastest way to lose an engineer's trust in the first five minutes of a call.

So the examples above are exactly what the instruments do, in the order they do it. The fastest way to test any of it is to hand us a file from your own programme and watch it open.

Next step

Use your own data as the case study.

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.