Which tests does an EVSE end-of-line station need?

A production test should link the manufactured unit to its configured tests and raw data. Choose the routine test scope from the approved product and manufacturing plan, then validate that the station can detect the intended defects.

1. Bind the unit to the recipe. Read the serial number, variant and firmware. Load the released recipe and verify that fixture, interface and instrument configuration match that variant.
2. Verify the station before the shift. Run the approved station check with a reference or controlled verification artefact. Confirm measurement status, interlocks and the treatment of expected failures.
3. Execute the defined sequence. Perform the approved routine safety checks and configured charging/signal/communication tests in the required order. Capture measured values, limits and individual verdicts.
4. Apply the failure route. Stop release on a failed mandatory check. Record the fault, rework identity and retest policy; retain the original failure alongside the retest.
5. Store and release. Save serial-linked recipe version, instrument status, raw measurements and overall verdict. Release the unit only after the mandatory record is complete.

Practical tip. Use a known-good unit to monitor repeatability, and a controlled known-fault artefact to test detection. A good-unit pass alone cannot demonstrate that the station catches the defect.

Turn the manufacturing requirement into a recipe

Assign a test recipe to the finished unit variant, not just to the production line. Include the hardware version, firmware, relevant interfaces and the checks required by the manufacturing acceptance plan. Record the source of each acceptance criterion. Development qualification and routine end-of-line checks serve different purposes: the production recipe should identify what is checked on this unit and which evidence comes from the approved design. Keep variant selection controlled so a successful execution of the wrong recipe is not accepted as a valid manufacturing record.

Make the unit identity part of every result

Connect the test result to the serial number or other agreed unit identifier, executed recipe and test-station configuration. Check that identity before storing the verdict. Retain the observations needed to explain a failed step, rather than saving only a final pass or fail. Define how incomplete executions, communication failures and operator interruptions are represented. A stopped sequence must remain distinguishable from a completed failed test. This makes the record useful for rework, traceability and later field diagnosis without relying on an operator's memory of what happened at the station.

Define the rework path before a unit fails

Agree who reviews a failure, which correction is allowed and which checks must be repeated after rework. Keep the failed attempt and the subsequent run connected to the same unit. Avoid overwriting the first record with a later pass. When a recipe or firmware changes, identify the affected variants and determine the required regression or qualification evidence. A stable production process needs both a repeatable test and a clear handling rule for exceptions. The test equipment records the executed behaviour; the manufacturing owner still defines the release and rework policy.

Example: charging communication fails on one unit

In an illustrative production case, the unit passes earlier selected checks but fails the defined charging-function sequence. Preserve the failed step and the relevant messages, then route the unit through the agreed rework process. After the correction, execute the required recipe and retain its link to the original failure. Check that the unit variant, firmware and test-station configuration are the intended ones in both records. The example describes traceability and exception handling; it does not prescribe a universal production test list, cycle time or acceptance threshold. Those decisions belong to the manufacturing requirements and the validated station configuration.