Product development
EVT, DVT and PVT: A Hardware Development Plan From Prototype to Mass Production
A hardware development plan is not a list of prototypes. It is a sequence of builds, each with a specific question to answer and a clear exit condition before the next build starts. The EVT, DVT and PVT gates give that sequence a shared vocabulary, so the design team, the tooling supplier and the factory all know what a build is supposed to prove and when the design is allowed to stop changing.
Before EVT: works-like and looks-like prototypes
Most products start with two separate early prototypes. A works-like prototype proves the function, often on development boards, off-the-shelf parts and a rough 3D printed housing. A looks-like prototype proves size, appearance and ergonomics with no working electronics inside. Keeping them separate at this stage is deliberate: it lets the team learn about function and form in parallel without waiting for either to be finished. The gate into EVT is the point where both are good enough to combine into one integrated design.
EVT (Engineering Validation Test): does the integrated design work?
EVT units combine the intended electronics with a near-final mechanical design for the first time. They are usually built with prototype methods such as CNC machining, 3D printing, urethane casting or soft tooling, and in small quantities, often tens of units. The goal is functional: do the board, enclosure, buttons, battery, thermal path and connectors all work together as designed? Expect EVT to surface fit problems, thermal issues and board revisions, and expect some teams to run a second EVT round before the design is ready to move on.
DVT (Design Validation Test): does the design meet its requirements?
DVT units are built with production-intent parts, which for plastic enclosures usually means the first samples from the actual injection mold rather than machined or printed housings. The question changes from "does it work" to "does it meet the specification": drop and environmental testing, reliability and cycle testing, cosmetic review of molded parts, and certification pre-scans for FCC, CE or UL where they apply. DVT is where the design should freeze. A change after DVT exit typically means modifying hardened steel, re-running tests and delaying certification, so the exit criteria should be written down before the build, not negotiated after it.
PVT (Production Validation Test): can the factory build it repeatably?
PVT is built on the real production line, with production tooling, assembly fixtures, test stations and operators, at a quantity large enough to measure yield. It validates the process rather than the design: cycle time, first-pass yield, test coverage, packaging and the documentation the factory will run from. PVT units are often sellable if they pass inspection, and PVT exit is normally the gate for starting mass production.
Where tooling decisions sit in the plan
The most expensive mistake in a hardware plan is releasing production steel while EVT issues are still open. A safer pattern is to finish EVT with the enclosure design stable, run a DFM review against the mold supplier's feedback, and then cut the tool that will produce DVT parts. If a launch date forces parts before the design is fully settled, a bridge or pilot mold can supply production-resin parts for DVT while leaving the production tool until the design has frozen.
What a useful plan document contains
For each gate, write down the question the build answers, the quantity, the build method, the test list, the exit criteria, the owner and the budget, and include a schedule buffer for at least one repeat round at EVT. Suppliers quote more accurately against this kind of plan, because they can see which parts are expected to change and which are expected to be production tooling.
Practical checklist
| Stage | Practical guidance | Why it matters |
|---|---|---|
| EVT | Integrated units from prototype methods (CNC, 3D print, casting, soft tooling), usually small quantities. | Proves the combined electronics and mechanical design actually works; expect board and fit revisions. |
| DVT | Production-intent parts, usually first samples from the production mold, plus reliability, drop and certification pre-testing. | Proves the design meets its specification and should end with a design freeze. |
| PVT | Built on the production line with production tooling, fixtures, test stations and operators, at yield-measuring quantity. | Proves the factory can build it repeatably; exit is the gate to mass production. |
| Tooling release | Cut production steel after EVT issues close and DFM feedback is resolved; use a bridge tool if parts are needed sooner. | Avoids paying for steel changes that the next build was always going to require. |
| Plan document | Question, quantity, build method, tests, exit criteria, owner, budget and buffer for each gate. | Gives suppliers a clear view of what will change and what is production-intent. |
Common mistakes to avoid
Approving visuals too early
A polished image can still hide structure, cost, tolerance, material or supplier risks.
Skipping useful outputs
Each stage should produce files, notes, samples or checklists that support the next decision.
Sharing sensitive files too soon
Start with safe context and move to detailed files after fit and NDA terms are clear.
Related questions
What is the difference between EVT, DVT and PVT?
EVT checks that the integrated design works, DVT checks that the production-intent design meets its specification and reliability requirements, and PVT checks that the factory can build it repeatably at acceptable yield.
When should production injection mold tooling be cut?
Usually after EVT issues are closed and a DFM review has been resolved with the mold supplier, so that DVT can be built from the production mold. Cutting steel before EVT closes is the most common source of costly tool changes.
Does a small startup need all three stages?
The names can be scaled down, but the questions cannot be skipped. A small team may combine builds or use fewer units, yet it still needs to prove the design works, meets its requirements, and can be built repeatably before committing to volume.
Related guides
Need help applying this to your product?
Send product type, stage and target market. We will reply with the next practical development path.