EV Charger Manufacturing Quality Plan

Incoming quality control

Critical components are verified against specification and, where the scheme requires it, against supplier certificates. Sampling is defined per part class rather than applied uniformly.

The reaction plan matters as much as the inspection. What happens to a failed lot, who is notified, and whether the line stops, are defined before the first failure rather than during it.

Process controls

• Torque control with calibrated tools and periodic verification
• Crimp control with pull testing to a defined sample rate
• ESD control that is measured, with records
• Wiring and routing verified against work instruction
• Sealing and gasket fitment confirmed before closure
• Programming with verified read-back
• Visual and workmanship criteria with reference samples

PFMEA and the control plan

The process failure mode and effects analysis identifies what can go wrong at each step and how severe it is. The control plan is the answer to that analysis: for each identified risk, which control catches it, at which step, against which limit.

A control plan that does not trace back to a PFMEA is a list of habits rather than a set of controls.

EOL and audit testing

Every unit passes end-of-line test. A defined sample additionally receives deeper audit testing, which catches problems the standard sequence is not designed to detect. Golden samples validate that the test systems themselves remain correct.

Traceability and records

Each unit links to its critical component lots, its process records, its test results and its firmware version. Retention period is defined and enforced.

The purpose is containment. When a component lot turns out to be bad, traceability is the difference between recalling forty units and recalling everything built that quarter.

Non-conformance, containment and CAPA

1. 1

   Raise

   Non-conformance recorded with evidence, not resolved informally at the station.

2. 2

   Contain

   Suspect stock quarantined, including work in progress and anything already shipped if applicable.

3. 3

   Analyse

   Root cause established rather than assumed.

4. 4

   Correct

   Immediate fix applied to affected units.

5. 5

   Prevent

   Process, control or design changed so the same cause cannot recur.

6. 6

   Verify

   Effectiveness confirmed after a defined period, and the action closed only then.

Supplier corrective action

Supplier issues follow the same discipline, with a defined response time, required root cause evidence and a change approval gate. Suppliers may not change process, material or sub-supplier without notification and approval.

Pilot gates and release to production

Start of production is released against quality evidence: pilot yield, closed non-conformances, verified control plan, calibration status and training records complete. It is a gate with criteria, not a date in a plan.

Two failure modes worth designing the plan around

A quality plan built from a standard covers what the standard anticipated. The useful additions come from failures an engineer has actually seen.

1. 1

   Internal cabling

   Cable quality and internal cable arrangement produce field failures that pass every acceptance test at end of line. We found this by analysing returns rather than by inspection, which is the point: the defect was not visible when the unit shipped. Treat cable as a controlled item with a named supplier and a defined routing, not as a commodity line.

2. 2

   Untested charging behaviour

   A line that tests only what can be checked without a vehicle will pass units whose charging sequence was never exercised. Control pilot behaviour, contactor sequencing and the metering path need a simulator on the line, or they are verified for the first time by a customer.

Both of these sit behind our field failure rate of roughly 0.5 percent, about one unit in two hundred. Neither was in the first version of any plan we wrote.