Key takeaways
- The hardware development life cycle runs from technical feasibility through requirements, schematics and PCB layout, firmware, enclosure, prototype validation, certification, and the design-for-manufacturing and production-handoff work that makes a product buildable at volume.
- Systems-engineering research treats this sequence as iterative, not linear: engineering validation and design validation findings routinely send a project back to an earlier stage.
- EVT, DVT and PVT are widely used hardware-industry terms for what each validation build must prove, not a formal engineering standard with one fixed definition.
- Cost and timeline both belong to separate questions. What matters here is the order of the stages and what each one has to deliver before the next one can start.
The hardware development life cycle is the sequence a physical product moves through from an idea to something that can be manufactured repeatably: technical feasibility, requirements and system architecture, electronic schematics and PCB layout, firmware, enclosure design, prototype build and validation, certification, and the design-for-manufacturing work that hands the product to a factory. Each stage has to produce a specific output before the next stage can rely on it, and real projects loop back between stages more often than a clean diagram suggests.
The stage map at a glance
| Stage | What it establishes | Typical output |
|---|---|---|
| 1. Technical feasibility | Whether the idea is technically achievable within budget | Risk list, scope, go/no-go |
| 2. Requirements and architecture | What the system must do and how its parts divide up | Requirements doc, block diagram |
| 3. Schematics and PCB layout | The electronic circuit and its physical board | Schematic, PCB files, BOM |
| 4. Firmware | How the electronics behave and respond | Source code, test builds |
| 5. Enclosure | The physical housing and fabrication method | CAD model, fabricated parts |
| 6. Prototype build and validation | Whether the integrated product performs as intended | EVT, DVT, PVT builds |
| 7. Certification | Whether the product meets regulatory requirements for its market | Test reports, approvals |
| 8. DFM and NPI | Whether the design can be built repeatably at volume | DFM report, production package |
Stage 1: Technical feasibility
Feasibility comes first because it is cheaper to discover a blocking problem on paper than after a schematic, enclosure or mold exists. It identifies the riskiest assumptions in the idea, such as whether a required component exists, whether a battery budget is realistic, or whether a mechanism can physically work in the space available, and turns an initial idea into a scoped, buildable plan. It is not a guarantee the product will work. It is a structured way to find out what still needs to be tested. See what product feasibility analysis actually covers for what this stage includes.
Stage 2: Requirements and system architecture
Once feasibility clears the idea, the project needs a written definition of what the product must do, under what conditions, and to what standard, before anyone starts designing circuits or firmware. System architecture is the decision about how those requirements divide between hardware, firmware and mechanical design, and which functions live on which subsystem.
- Requirements should state the essential function, target users, intended markets and acceptance criteria, not just a feature list.
- Architecture decisions made here, such as whether to use an off-the-shelf wireless module or a custom radio, affect every stage that follows.
- In our experience at Inventornest, clients do not need to arrive with a complete engineering specification. Feasibility and early architecture work exist to turn a rough idea into defined requirements and a realistic scope.
In our experience at Inventornest, the choice between a custom PCB and an off-the-shelf module usually comes down to whether the module can meet the required dimensions, power consumption, interfaces and production cost. A low-volume product with tight space constraints may need a custom board even though a higher-volume product is well served by a module. Volume alone does not decide it.
Stage 3: Electronic schematics and PCB layout
The schematic defines the circuit; the PCB layout turns that circuit into a physical board that can actually be fabricated and assembled. Board design is governed by generic design standards such as IPC-2221, published by the Global Electronics Association (formerly IPC, which took the Global Electronics Association name in 2025 while keeping its existing standard numbers), which sets the baseline requirements for printed board design across single-sided, double-sided and multilayer boards. The data handoff from design to the fabricator and assembler, covering the files, bill of materials and manufacturing data a factory actually needs, is addressed separately by IPC-2581.
Two PCB development iterations are typical for a first prototype in our experience at Inventornest. Depending on complexity and the issues found, a project may need four or five. In our experience at Inventornest, revisions can involve both PCB changes and firmware updates together, when that work falls within the agreed project scope. For what happens on the fabrication and assembly side specifically, see PCB fabrication and assembly.
Stage 4: Firmware development
Firmware is what makes the electronics behave: reading sensors, driving outputs, managing power states and responding to the outside world. It is usually scoped as part of the broader software life cycle rather than treated as its own standard. The embedded product development life cycle shows how the firmware and hardware stages run side by side. The current ISO/IEC/IEEE 12207 systems-and-software life-cycle standard, updated in 2026, frames this work as implementation and integration activities inside a larger process that also covers architecture definition, verification and operations, without mandating any single development methodology.
In our experience at Inventornest, firmware and PCB design develop together: we develop the PCB before finalizing the enclosure, so its dimensions and layout inform the enclosure design, and firmware changes can accompany PCB revisions when a substitution or design change affects how a component is driven.
Stage 5: Enclosure design and fabrication method
The enclosure stage decides the physical housing and how it will be made: 3D printing, CNC machining or sheet metal fabrication for prototypes, and injection molding once a design is validated and the planned quantity justifies the tooling cost. 3D printing is normally used during development because it allows changes without committing to tooling. There is no fixed unit threshold for switching to molding; the decision depends on enclosure complexity, material, finish, tolerances and tooling cost against the savings per unit at the planned volume. See how 3D printing changed product development and low-volume versus mass production for the fuller picture on when each method applies.
Battery and enclosure decisions are linked directly. In our experience at Inventornest, increasing battery size to meet a runtime requirement found during testing requires additional internal space, which can force enclosure changes late if battery sizing was not settled early alongside the PCB.
Stage 6: Prototype build and validation
This is where the integrated product gets tested as a whole rather than as separate subsystems. Hardware teams commonly describe this work using the terms engineering validation test (EVT), design validation test (DVT) and production validation test (PVT), though these are industry convention rather than a formal standards-body definition: no ISO, IEC or IPC standard formally codifies EVT, DVT or PVT. In practice, each is understood to test a different kind of risk, moving from “does the engineering work” toward “does the production process reproduce it reliably.” For where each stage sits and what it is meant to prove, see prototype vs. MVP: what’s the difference, and for the standards and thresholds a prototype should actually be tested against, see the prototype testing checklist.
In our experience at Inventornest, common products with established functions and few unique features can pass the first build without changes. When issues do occur, one recurring cause is a component performing differently in the real application than its supplier’s specification claimed.
Stage 7: Certification
Certification requirements depend heavily on the target market and whether the product transmits radio signals. In the US, the FCC regulates both intentional radiators, devices that intentionally generate and emit RF energy such as Wi-Fi or Bluetooth radios, and unintentional radiators, devices whose internal digital circuitry generates RF energy without being designed to transmit it, under 47 CFR Part 15. Under current FCC rules, a manufacturer uses one of two paths to equipment authorization: Supplier’s Declaration of Conformity, used for most unintentional radiators, where the product is tested at a qualified lab and the responsible party self-declares compliance, or full Certification, required for intentional radiators such as Wi-Fi or Bluetooth transmitters, which requires testing at an FCC-recognized accredited lab and review by a Telecommunication Certification Body before the FCC issues a grant. FCC rules are a US-only regime: CE marking in the EU and UL safety certification in North America are separate systems with separate requirements, and a product sold in multiple markets may need more than one. See how FCC-recognized test labs fit into that process for the testing side of this stage.
Stage 8: DFM and NPI: preparing for the production handoff
Design for manufacturing (DFM) review checks whether a design that works as a prototype can actually be built repeatedly, economically and at the required quality on a production line, covering land patterns, solder mask and assembly acceptability against the Global Electronics Association’s board-design and assembly standards. New product introduction (NPI), by contrast, is the narrower, later stage: per the Product Development and Management Association’s glossary, the launch or commercialization of a new product into the marketplace, taking place at the end of a successful product development project. The two terms get used loosely, but NPD describes the whole development process while NPI specifically describes the production ramp-up and handoff at the end of it.
Formal manufacturing preparation and the specific checks performed depend on the agreed scope and the product. See why DFM timing locks in certain decisions early, NPD vs. NPI, and what happens during a first production run for the pilot batch that actually proves the process repeats. Quality checks that continue after launch, such as AQL sampling and first article inspection, are covered in quality control in manufacturing.
Why the stages loop back
The eight-stage map above describes a typical sequence, not a one-way pipeline. The Systems Engineering Body of Knowledge (SEBoK), a reference maintained with INCOSE, treats iteration as a normal part of the life cycle: requirements and architecture decisions commonly get revisited once a prototype or validation build surfaces a problem, and engineering review milestones are explicitly built around these loops rather than a straight line through the stages. A battery-space constraint discovered in DVT, for example, can send a project back to enclosure or even requirements, and that is the life cycle working as intended, not a failure of planning.
Frequently asked questions
What are the main stages of hardware product development?
Technical feasibility, requirements and architecture, schematics and PCB layout, firmware, enclosure design, prototype build and validation, certification, and design for manufacturing and production handoff.
What is the difference between EVT, DVT and PVT?
They describe different validation objectives rather than a fixed number of prototype units: broadly, EVT checks whether the engineering works, DVT checks it against the full set of product and interface requirements, and PVT checks whether the production process itself reproduces the result reliably.
What is the difference between NPD and NPI?
NPD covers the entire development process from strategy through commercialization. NPI is the narrower launch and market-introduction stage that happens after development succeeds.
Does hardware development always move in a straight line through these stages?
No. Findings during validation testing routinely send a project back to an earlier stage, such as architecture or enclosure design, and systems-engineering practice treats that as a normal part of the life cycle.
When does certification happen in the process?
Certification requirements should be considered during requirements and architecture, since they affect design choices, but formal testing typically happens once a design-validated build exists and before production handoff.
What does design for manufacturing actually check?
Whether the validated design can be built repeatedly, at the required quality, using the intended production components and process, rather than only assembled once as a prototype.
Where Inventornest fits
Inventornest handles the design and development stages of this life cycle in-house: technical feasibility, electronic schematics, firmware, PCB development, and enclosure design and fabrication by 3D printing, sheet metal or CNC machining, followed by assembly, testing and iteration. Every stage goes through internal technical review before the client sees it. For production, we retain responsibility for engineering coordination and design clarification when manufacturing is handled by our manufacturing partners, and external laboratory certification and patent-attorney services remain separate professional activities we coordinate around. If you need expert input at a single stage, or your project is already under way, product development consulting covers any point in this life cycle. If you want help mapping where your own product sits in it, get a quote.
