EV Charger Pilot Production and SOP Readiness

Engineering build

A small quantity built by engineering, against the documentation the line will use rather than against knowledge held in someone's head. The output is a corrected document set as much as it is hardware.

If engineering cannot build the product from the written instructions, neither will production, and that finding is worth more than the units.

Controlled pilot

Built by production operators, at low rate, with engineering present to observe rather than to assist. What operators do differently from the instruction is data, not error, and it usually indicates an instruction that does not match the physical reality of the assembly.

Line qualification

Fixtures, torque tools and test stations demonstrate repeatability across operators and across shifts. A station that performs on the day shift with its designer present has not been qualified.

Rate trial

The line holds takt and yield over a sustained run. This is where bottlenecks that were invisible at pilot rate appear, and where the overall equipment effectiveness assumption in the capacity model gets tested against reality.

Release criteria

Start of production is a gate with conditions, not a date in a plan. Typical conditions include closed non-conformances from the pilot, a verified control plan, calibration current on every station, complete training records, and yield sustained above an agreed level during the rate trial.

The specific thresholds belong in your quality plan rather than on this page, because they depend on the product, the scheme and the customer.

The stabilization window

After release, an agreed period during which yield, the defect Pareto and early field returns are reviewed jointly before responsibility fully transfers. Problems that depend on volume, operator rotation and supplier lot variation become visible here.

Some of those are detectable earlier through rate trials and audit sampling. Stabilization is the cheapest place to find them because responsibility has not yet fully moved, and both parties still have an interest in the answer.

What a pilot should produce besides units

• A corrected and re-released document set
• A closed issue list with root causes, not a list of observations
• Verified test limits, adjusted where characterisation was wrong
• Confirmed cycle times per station, feeding the capacity model
• Trained and signed-off operators on every station
• A defect Pareto that tells you where to spend improvement effort

What a pilot is actually for

A pilot run is not a small production run. It is the point at which the assumptions in the transfer package meet a real line, real operators and real parts, and some of them turn out to be wrong.

On our European programme the pilot stage is where the supply base problems surfaced properly. Components that had been qualified on paper behaved differently in volume, and the second sources that had looked adequate were not all adequate. That is the pilot doing its job. Finding it at production scale would have been considerably more expensive.

Expect the pilot to change things. A pilot that changes nothing was either unnecessary or was not looked at hard enough.

• The bill of materials, where a qualified part turns out to behave differently in volume
• The work instructions, where a step that reads clearly is performed three different ways by three operators
• The test limits, where the line finds a failure mode the design team had not anticipated