EVT, DVT, PVT Stage Gate Definitions

This EVT, DVT, PVT article is one of the first pieces of content I wrote for Instrumental in 2015. At the time, there was very little on the internet about the lives of hardware engineers, our struggles, and how to excel at our jobs. This piece rapidly became the most back-linked, downloaded and referenced content on instrumental.com, with leaders telling me they embedded it in their onboarding, training, and company wikis (we can see the referring domains as proof) -- it's been referenced tens of thousands of times. In the intervening years, this knowledge became more widespread and adopted by many of the world's best electronics companies (not to mention copycat articles on the web). Many years later, this article inspired the Build Better Handbook as part of our mission to improve how teams build. I am reposting it here as foundational content to assist leaders in clearly defining the stage gates of their own development process.

While software engineers may write code, hardware engineers speak it. We love our three-letter acronyms (TLAs): EVT, DVT, PVT, MP; OK, NG, FA, CA; PD, EPM, OPM; PRD, DRP, BOM. Goodness, how do we even understand each other?

I spend a lot of time speaking with engineering and operations leaders at hardware companies about their manufacturing pain points. During those conversations, we always discuss those pain points in the context of product maturity and the builds: EVT, DVT, PVT and MP. While a few companies use slightly different nomenclature, the basic structure is consistent: you build prototypes of the design multiple times to zero in on the final, production-ready design. However, I've discovered differences in understanding across the electronics industry on what EVT and DVT product maturity even mean -- often even within the same team.

I’ve put together this guide to explain the builds — cobbled together from my years of experience as a Product Design Engineer, visibility into over a hundred NPI programs, and hundreds of interviews with engineers and their leaders.

Proto

The Proto build is a small test run of key product concepts to gain confidence that they can work — potentially a combination of different form factors including looks-like and works-like.

Purpose: to understand risks around specific modules or designs, usually with multiple variants in low quantities, such as:

1. Fragility of coverglass in drop test with different adhesives, perhaps done on dummy housing bucks.
2. Waterproofness of five different button seal designs.

Typical Quantities: 10 or fewer, sometimes no “full systems” are even built.

1. Parts may be “stand-ins” or rapidly prototyped (which may change results for better or worse).
2. Sub-modules do not have to be integrated — units may be “works like” or “looks like.”

Things that Go Wrong:

1. Part quality is poor, resulting in incorrect dimensions or an interference was missed in the CAD (3D model), so parts do not fit together and have to be modified by hand.
2. Pin 1s on connectors were not correctly mapped, so things do not electrically work even when plugged together.
3. The intended design fails miserably during testing and needs to be redesigned.

Exit Criteria: one design concept for the product that the team has reasonable confidence is three major iterations or less from a mass-production worthy design.

EVT (Engineering Validation Test)

The EVT build is the first time you combine looks-like and works-like into one form factor, with production intent materials and manufacturing processes.

Purpose:

1. To select the production intent design, sometimes from a build matrix of options.
2. To identify all of the issues that need to be fixed with that design.

Typical Quantities: 100 to 1,000.

1. Units must be fully functional and testable, made from the intended materials and with the intended manufacturing process, but may be from soft-tools (if you’re using 3D printed parts, it’s not EVT!).
2. All functional test stations must be present and collecting data.

Things that Go Wrong:

1. A new revision of an intended design does not work after reliability testing.
2. Tighter than expected (or capable) tolerances are needed to meet the intended performance specifications — such as with an antenna element.
3. Depending on product complexity, up to ~40% of the units built may fail for a variety of functional or performance reasons and need to be analyzed.
4. Engineering has started the battle to get glue processes, hand-soldering, environmental seals, and other tricky steps under control.

Exit Criteria: one production-worthy configuration that meets all of the product requirements for functionality, performance and reliability.

DVT (Design Validation Test)

The DVT build is supposed to be one configuration of your production-worthy design, made of components from production processes (and hard tools) and on a line following production procedures. I believe very few companies stick to this requirement — because even if miraculously there are no outstanding issues, there may be parallel efforts to cut costs or increase yields that create additional configurations to build.

If you do have functional, performance or reliability issues that are driving Plan B and Plan C configurations at this stage, it can be costly because each of those alternates needs to be built in “full quantity” to ensure that design can be fully mass-production qualified by the end of the build. I believe that’s the real test for whether you are at DVT: if you are running side configurations of 20 units, you are fooling yourself and should call it EVT2.

Purpose:

1. To verify mass production yields with one production-worthy design (one configuration for each shipping SKU).
2. To qualify the first hard tool for every part in the assembly.

Typical Quantities: 300 to 2,000.

1. All parts should be from hard tools or mass-production-capable processes.
2. All functional test stations must be present with limits in place to understand true yields.

Things that Go Wrong:

1. High functional fallout rates — requiring the need for fast failure analysis and corrective actions.
2. Cosmetic yields are 0% — there may be an effort to try to track down and fix cosmetic aggressors, but it is usually fruitless because your cosmetic part suppliers are likely still shipping scratched parts (and you have to waive them).
3. At least one key process, such as gluing, is still not under control —
   1. DOEs are run with alternate glues or curing parameters.
   2. There are nightly calls with vendors demanding support or giving updates to hardware company executives.

Exit Criteria: high confidence in all corrective actions for any issue that causes unacceptable unit yields using mass production parts made from mass production tools.

PVT (Production Validation Test)

PVT is the “last build” — the units you are building are supposedly intended to be sold to customers, if they pass all of your test stations. PVT typically transitions directly into Ramp and Mass Production or a Pilot build with no time gap.

Purpose: to verify mass production yields at mass production speeds.

1. Validate and qualify additional tools needed to support quantities for early ramp.
2. No parallel experimental units allowed. (I have never seen this happen, but it is a goal that should be driven to for as long as possible.)

Typical Quantities: 1K to 20K

1. All units are intended to be sold to customers
2. The build is potentially phased — red, yellow, green is common — indicating “maturity” of the production process, which includes a combination of operator training level, line speed, and line yield.

Things that Go Wrong:

1. There is almost always at least one issue that is still outstanding at the start of PVT — this is likely the item at highest risk of impacting your schedule.
2. There is usually at least one vendor whose yields are way lower than expected, and because they cannot produce at the quantities promised, input is gated by their deliveries.
3. If you have a high cosmetic standard, your cosmetic yield likely starts at 0%. Unless you decide to loosen your standard, the conventional way to improve it is to knowingly input units to a 0% yield line and painstakingly seek places where damage occurs and improve them. This process can take weeks and hundreds or thousands of units. An Instrumental AI-based visual inspection system can streamline and significantly accelerate this process.

Exit Criteria: mass production yields at mass production speeds on at least one line, and replication to other lines (if applicable) has started.

Ramp and MP (Mass Production)

PVT flows immediately into the phase of the program called Ramp, where parallel assembly lines are being brought up to increase daily output volume. Mass Production is a superset of Ramp and the sustaining production that follows.

Purpose:

1. Bring up multiple lines in parallel to support high volume.
2. Continue to improve ongoing yield.
3. Qualify additional tools or vendors.
4. Make design changes based on returns, Early Field Failure Analysis (EFFA), or cost-down efforts.

Things that Go Wrong:

1. Vendors change processing parameters or take down tools for maintenance, resulting in dimensional or quality shifts that can cause line failures.
2. Parts from unqualified tools are allowed on the line and cause failures.
3. A single-sourced part becomes the supply gate, usually due to ongoing yield issues.
4. Quality tends to decrease as engineering is pulled away and the factory is left unsupervised.

Beware XVT: where X stands for 🤞

Timing for the build process outlined above is driven by the need to iterate hardware to get the design right. That need often comes into direct conflict with the realities of the market: if you’re building a toy, for example, it must be ready to ship for Christmas. This tension between the iteration process and the market-driven schedule can do weird and sometimes dangerous things to the development process. While there’s much to discuss, I wanted to end with a cautionary note about the nuclear option: XVT.

XVT is a fabrication of over-optimistic program managers and operations executives who believe that it’s possible to enter a build with EVT parts and complete DVT exit criteria (the X being a stand-in for either an E or a D, where everyone crosses their fingers that by the end it’s a D). XVT doesn’t stand for anything, but if it did, it would be No Validation Test. My experience is that investing massive DVT-scale resources into an EVT maturity design does not get your product out faster.