What PVT proves
By PVT the design should be frozen. DVT showed that the product meets its requirements; PVT shows that the factory can make it the same way every time. The questions shift from engineering to manufacturing. Is yield where it needs to be? Does each station meet its cycle time? Do the tests catch real defects? Can operators follow the work instructions as written?
Changes during PVT should be process changes: fixture adjustments, line balancing, updated work instructions, tuned test limits. A design change found necessary in PVT means DVT missed something, and it may require repeating some DVT tests. For how PVT relates to EVT and DVT, see the EVT, DVT and PVT guide.
What a PVT build looks like
- Line: the stations planned for mass production, not an NPI or engineering area.
- People: production operators trained from the work instructions, not engineers.
- Rate: long enough at or near the planned rate to expose bottlenecks. A handful of units an hour on a line designed for many more hides most problems.
- Quantity: often hundreds to a few thousand units, depending on line rate and product. Size it so every station and shift produces meaningful yield data.
- Materials: bought through the normal supply chain and checked at normal incoming inspection (IQC), so sourcing problems surface now.
- Name: many factories call this build the pilot run. Some use that name for a smaller line trial before PVT, so confirm which build is meant.
PVT units are often sold if they pass outgoing inspection and the brand accepts them. Agree this in advance, along with what happens to units that do not pass.
What to measure during PVT
| Metric | What it tells you |
|---|---|
| First pass yield at each test station | Where defects are created and whether the process is capable |
| Final yield and scrap | The cost of quality at volume |
| Cycle time per station against takt time | Whether the line can meet planned output, and where the bottleneck is |
| Defect Pareto | The few causes behind most failures |
| Test escapes | Defects found at outgoing inspection that line tests should have caught |
| Cpk on critical dimensions or parameters | Whether the process is centered and stable enough |
| Rework rate | Hidden cost and risk; track reworked units separately |
PVT exit criteria
- Yield meets a target agreed before the build and defined precisely (first pass or final, per station or for the line, with or without rework), and the top remaining defects have corrective actions in progress
- Every station meets its cycle time, or the capacity plan accounts for the shortfall
- Test coverage is confirmed, and test escapes found in PVT are closed
- Work instructions, control plan and inspection criteria are updated with what PVT taught
- Outgoing inspection and sampling to AQL run as they will in production
- Packaging, labels and shipping documents are validated
- The brand signs off the production golden sample and the release to mass production
Common PVT mistakes
- Running PVT on the NPI line with engineers. Yield looks good and says nothing about production.
- Releasing on one good shift. A clean shift can hide problems that appear with a second shift, a new material lot or a full day at rate. Judge the trend across the whole build.
- Treating PVT as a formality. When schedules slip, PVT is often cut to a short run, and the problems it would have found appear during mass production, where they cost more.
- Mixing design and process changes. If the design changes during PVT, record which units are affected and repeat the relevant validation.
- Building from hand-carried parts. If PVT runs on components expedited by engineers or bought from distributors, the normal supply chain and incoming inspection have not been tested.
Where Zefinity helps
Zefinity’s Launch service supports PVT with your OEM/ODM partner through production readiness reviews, yield and defect follow-up and corrective action tracking up to the release decision. See Launch: navigate NPI and scale production and the pilot production run guide.
Frequently asked questions
Does PVT need new certification testing?
Normally not, if PVT units match the design that was certified, because PVT is meant to change the process rather than the product. Certification applies to the product as tested, though. If a fix made during PVT touches the radio, antenna, enclosure, power circuit or battery, ask the test lab whether retesting or an update to the filing is needed before you ship.
Who decides that PVT has passed?
The brand, on evidence from the factory. The factory reports yield, cycle time, test coverage and inspection results against exit criteria agreed before the build, and the brand reviews them and signs the release to mass production. The factory has a natural interest in starting volume orders, while the brand carries the warranty and market risk, so the decision should not default to the supplier's schedule.