Founder guide
Hardware Product Development Roadmap: From Idea to Mass Production (2026)
An electronic product usually goes from idea to mass production through six stages — feasibility, a works-like prototype, engineering validation (EVT), design validation (DVT) with certification, production validation (PVT) with a pilot run, and the production ramp — which takes about 9–18 months for a connected consumer product. For planning, budget roughly $30,000–$150,000+ in engineering with an India-based team ($50,000–$250,000+ with a US/EU firm), plus tooling, certification lab fees and first inventory, which are paid separately.
Key facts
| Question | Short answer |
|---|---|
| Stages | Feasibility → proof of concept → EVT → DVT → PVT → mass production |
| Typical calendar time | 9–18 months for a connected consumer product; 12–24+ months for a regulated medical device |
| Engineering budget (planning) | $30,000–$150,000+ India-based team; $50,000–$250,000+ US/EU firm; medical $120,000–$400,000+ |
| Paid separately | Enclosure tooling ($3,000–$20,000+), test fixtures ($2,000–$10,000), certification lab fees, first inventory |
| Pilot run | Typically 50–500 units built on the production line during PVT |
| Founder’s job | Own the requirements, target price and BOM cost, markets, budget gates and the contracts |
| IP | Mutual NDA first; foreground IP assigned to you; all native source files listed as deliverables |
The ranges on this page are planning figures from Rapid Circuitry’s project experience since 2020, stated in US dollars at blended engineering rates. They are consistent with our hardware development cost benchmarks and the IoT development cost guide. They are not quotes: your requirements, the number of design iterations and the markets you certify for move every number.
The six stages at a glance
The industry names below (EVT, DVT, PVT) come from consumer-electronics manufacturing, but every serious hardware programme — medical, industrial or consumer — follows the same logic: prove the idea, prove the design, prove the factory. Each stage produces something you can inspect and ends with a gate where you decide whether to spend the next stage’s budget.
| Stage | What it produces | Typical duration | Engineering cost (India-based team) | Units built |
|---|---|---|---|---|
| 1. Feasibility & requirements | Written requirements (PRD), architecture, risk list, target BOM cost, certification plan | 2–6 weeks | $3,000–$10,000 | 0–3 dev-kit rigs |
| 2. Proof of concept (works-like) | Riskiest functions working on dev kits or a first custom board; demo firmware and app | 6–12 weeks | $8,000–$30,000 | 1–5 |
| 3. EVT — engineering validation | Custom PCBs close to final form, full schematic/layout, firmware feature-complete for test, bench test report | 6–12 weeks | $12,000–$50,000 | 10–50 |
| 4. DVT — design validation | Production-intent design, tooled enclosure (T1 samples), reliability tests, pre-compliance and formal certification | 8–14 weeks | $8,000–$40,000 + lab fees | 50–200 |
| 5. PVT — production validation | Pilot run on the real line with production fixtures, test software and packaging; yield data | 4–8 weeks | $4,000–$20,000 + unit cost | 50–500 |
| 6. Mass production ramp | Stable yield, sustaining engineering, second sources, field-return loop | 4–8 weeks to stabilise | Retainer or per change | 1,000+ |
What happens in each stage
1. Feasibility and requirements (2–6 weeks)
The most valuable few weeks of the programme. The team turns your idea into a written product requirements document (PRD): what the product does, who uses it and where, battery life, size, connectivity, target retail price, target bill-of-materials (BOM) cost, markets and certifications, and launch volume. Engineers then propose an architecture — which microcontroller or radio module, which sensors, how it charges, what the app and cloud must do — and list the technical risks. The riskiest item (a sensor that must work through fabric, a battery target that may be impossible) is tested on a bench before anyone designs a PCB.
Exit gate: a signed-off PRD, an architecture you understand, a BOM cost estimate at your launch volume, and no unresolved “this might not be physically possible” risks.
2. Proof of concept — the works-like prototype (6–12 weeks)
A prototype that works like the product but does not yet look like it. It may be built from development kits and wires, or on a first custom board. Firmware demonstrates the core functions; the app shows the main screens. This is what you show investors or use for early user interviews. Do not freeze the industrial design yet — the electronics will still change size.
Exit gate: the core function is proven with real sensors and a real radio, and power consumption measurements confirm the battery target is achievable.
3. EVT — engineering validation (6–12 weeks)
The electronics are designed properly: complete schematics, a PCB layout in (or close to) the final shape, a BOM with manufacturer part numbers, and firmware that implements every feature needed for testing. The team builds 10–50 boards, often in 3D-printed or CNC housings, and runs bench tests against the PRD: power rails, radio range, battery life, sensor accuracy, charging, temperature. Expect to find bugs; that is the purpose of EVT. One PCB respin between EVT and DVT is normal and should be in your plan.
Exit gate: every requirement has a measured result, and open issues are understood with a fix planned for the DVT build.
4. DVT — design validation (8–14 weeks)
The design is now production intent: the enclosure comes from the first injection-mould shots (T1 samples), the PCB includes DVT fixes, and firmware is close to release. DVT units go through reliability tests (drop, thermal cycling, humidity, button and connector life, battery safety) and certification. Run pre-compliance EMC and radio scans first — a failed formal lab session costs a re-test fee plus weeks. Formal testing covers the markets in your PRD, for example FCC in the US, CE under the Radio Equipment Directive in the EU, and BIS/WPC in India (see our EU and US market-access guide and India certification guide).
Exit gate: certification test reports passed (or remaining actions scheduled), reliability results within limits, and a frozen design released to the factory.
5. PVT — production validation and the pilot run (4–8 weeks)
The contract manufacturer (EMS) builds units on the real line with production stencils, fixtures, programming jigs and end-of-line test software. The questions change from “does it work?” to “can the factory build it consistently?”: first-pass yield, test coverage, cycle time, packaging and labelling. PVT units are normally sellable and often go to beta customers or early backers.
6. Mass production ramp
Volume increases in steps while engineering watches yield and field returns. Expect sustaining work: a component goes end-of-life, a supplier changes a part, a firmware update fixes a field issue. Budget a small retainer or a per-change arrangement with your design partner for the first year.
Decisions a non-technical founder must own
You do not need to be an engineer to run a hardware programme, but some decisions cannot be delegated because they set the budget and the schedule. Write your answers down before the first engineering call; a good design partner will challenge them, which is useful.
- Who buys it and at what price. Retail price constrains BOM cost. A common consumer-hardware rule of thumb is that retail price lands at roughly 3–5× the landed product cost once channel margin, returns and marketing are included.
- Must-have vs. version-two features. Every feature adds firmware, testing and sometimes certification. Launching with fewer features is the most reliable way to protect the schedule.
- Markets. Each country adds certifications and labelling. Selling in the US, EU and India from day one roughly triples the certification work compared with one market.
- Battery life, size and charging. These three trade off against each other and against cost. Decide which one wins.
- Launch volume and channel. 500 units on Kickstarter and 20,000 units in retail lead to different designs (pre-certified modules and simple enclosures versus custom radios and multi-cavity moulds).
- Budget gates. Release money stage by stage against exit criteria rather than committing the whole budget up front.
- Who owns the app and cloud. They are software products with their own running costs; decide early whether the design partner builds them or your own team does.
Choosing and managing a design partner
The biggest risk for a first-time founder is not a technical failure; it is losing months to hand-offs between separate freelancers for electronics, firmware, enclosure and app. Whoever you hire, one party should be accountable for the product working as a whole. Our partner-selection guide has a weighted scorecard and 20 due-diligence questions; the short version for founders:
- Ask to speak with the engineers who will do the work, and ask them to walk you through a comparable past product.
- Ask for a fixed price per stage after a paid feasibility or requirements phase, with written exit criteria for each stage.
- Get access to the design repository and issue tracker from week one, not at the end.
- Agree a weekly rhythm: a short written status, a demo whenever there is something to show, and an updated risk list.
- Make sure prototypes are shipped to you at each stage so you can hold what you are paying for.
- Check how change requests are priced before you need one.
Red flags: a quote without a requirements phase, no mention of certification or DFM (design for manufacturing), reluctance to name who owns the source files, and schedules that skip EVT “to save time”. Our DFM checklist shows what a manufacturing-ready design review covers.
IP ownership, NDAs and the files you must receive
Founders often worry that a design firm will steal the idea. In practice the bigger risk is ending the project without the files you need to manufacture, fix or move the product. Three documents protect you:
- Mutual NDA signed before you share details beyond the one-line pitch.
- Development agreement with IP assignment: all project-specific (foreground) IP — designs, firmware, app code, documentation — assigned to your company on payment; a perpetual, royalty-free licence to any pre-existing (background) IP or libraries embedded in your product.
- A deliverables list attached to the contract, so “all design files” is not open to interpretation.
| Deliverable | Format | Why it matters |
|---|---|---|
| Schematic | Native project (KiCad, Altium…) + PDF | Needed to modify or debug the design with any engineer |
| PCB layout | Native project + stack-up and fab notes | Gerbers alone cannot be edited |
| Manufacturing outputs | Gerber/ODB++/IPC-2581, drill, pick-and-place, assembly drawings | What the factory builds from |
| BOM | Spreadsheet with manufacturer part numbers and approved alternates | Lets any EMS quote and source parts |
| Firmware | Full repository with history, build instructions, toolchain and SDK versions | A binary cannot be maintained or updated |
| App and cloud | Source repositories, infrastructure configuration, account ownership | Your product stops working if these accounts are not yours |
| Mechanical CAD | STEP + native files, drawings | Needed for tooling changes |
| Test | Test specification, fixture design, end-of-line test software | Needed to change factories |
If your idea may be patentable, speak to a patent attorney before you publish anything, launch a crowdfunding page or show it at a trade show; public disclosure can destroy novelty in many jurisdictions (WIPO). In the US, a provisional patent application gives a 12-month window at modest cost. This section is general information, not legal advice.
Pilot runs: why 50–500 units
A pilot run is the first build on the production line, usually during PVT. It is large enough to show the factory’s real yield and the failure modes that never appear in 10 hand-built prototypes, and small enough that a problem does not strand a warehouse of stock.
| Pilot size | Typical use | What you learn |
|---|---|---|
| 50–100 units | Regulated or high-cost products, B2B trials | Assembly issues, test-fixture coverage, first yield figure |
| 100–300 units | Consumer beta, crowdfunding early backers | Yield by failure type, packaging, unboxing, app onboarding at scale |
| 300–500 units | Retail launch preparation | Line balance, cycle time, a statistically meaningful yield, early field returns |
Unit cost at pilot volume is much higher than at mass production. Our IoT cost guide’s planning figures put an assembled PCB at roughly $20–$80 per unit at 100–500 units versus $8–$25 at 1,000–5,000 units for a typical IoT board. Negotiate with the EMS so that stencils and fixtures made for the pilot are reused in production.
Example plan: a Bluetooth sensor with an app
A planning scenario, not a past project: a battery-powered wearable sensor with a pre-certified Bluetooth Low Energy module, one four-layer PCB, a rechargeable battery, an iOS/Android app and launch in the US and EU.
| Months | Stage | Key outputs |
|---|---|---|
| 0–1 | Feasibility | PRD, architecture, BOM estimate, sensor bench test |
| 1–3 | Proof of concept | Dev-kit prototype, power measurements, app wireframes |
| 3–6 | EVT | Custom PCB rev A, firmware features, app beta, industrial design frozen |
| 6–9 | DVT | PCB rev B, T1 enclosure, reliability and pre-compliance, FCC and CE testing |
| 9–11 | PVT | Pilot run of 200 units, end-of-line test, packaging |
| 11–12+ | Ramp | First production order; sustaining engineering |
Common mistakes that add months
- Freezing the industrial design before the electronics are sized — the board then does not fit.
- Treating certification as a final step instead of a design input (antenna placement, enclosure material and cable routing decide EMC results).
- Choosing parts without checking lifecycle status and lead time.
- Skipping DFM review, so the first factory build reveals unassemblable footprints or untestable boards.
- Not owning cloud, app-store and domain accounts.
- Under-budgeting the app: for many connected products it is a third of the engineering effort.
- No contingency. Add 15–25% to whatever the plan says.
If a project is already off the rails, our guide on rescuing a stalled hardware project covers audits and vendor takeovers.
How Rapid Circuitry runs these stages
Rapid Circuitry is an electronics design and engineering firm in Hyderabad, India, founded in 2020, with a team of 25+ engineers and 500+ projects delivered. We cover electronics design, PCB design, firmware, mobile apps and cloud under one team, with prototyping and manufacturing support through pilot builds. We quote a fixed price per stage after a requirements phase and assign all project IP and source files to the client. For an example of a consumer product delivered end to end, see the multi-connectivity pet tracker case study.
Frequently asked questions
What are the actual steps to go from a sketch to a manufactured electronic product?
Most electronic products go through six stages: feasibility and requirements (2–6 weeks), a works-like proof-of-concept prototype (6–12 weeks), engineering validation or EVT (6–12 weeks), design validation or DVT including certification testing (8–14 weeks), production validation or PVT with a pilot run of roughly 50–500 units (4–8 weeks), and the mass-production ramp. Each stage ends with a go/no-go review against written exit criteria, so you only commit the next stage’s budget when the previous stage has proved what it was meant to prove.
Who do I need to hire at each step of hardware product development?
You need an electronics engineer (schematic and PCB), a firmware engineer, an industrial or mechanical designer for the enclosure, and usually a mobile-app or cloud developer. From DVT onward you also need a test lab for certification and a contract manufacturer (EMS) for assembly. A non-technical founder can either hire and coordinate these specialists separately or appoint one design partner that covers electronics, firmware and app and manages the lab and factory hand-offs; the second option costs more per hour but removes most of the coordination risk.
How much does it cost to develop a hardware product from idea to prototype?
As a planning range, a first works-like prototype of a connected product costs about $10,000–$40,000 in engineering with an India-based team, including feasibility work. Taking the same product all the way to a production-ready design typically adds up to $30,000–$150,000+ with an India-based team or $50,000–$250,000+ with a US/EU firm. Enclosure tooling, certification lab fees and the first production inventory are budgeted separately.
How long does it take to get from idea to a first working prototype for a Bluetooth gadget with a phone app?
Plan about 8–16 weeks: 2–4 weeks to write the requirements and choose the architecture, then 6–12 weeks to build a works-like prototype on a development kit or a first custom board with basic firmware and a simple app. A looks-like and works-like prototype in the final form factor is an EVT deliverable and typically arrives 3–6 months after kickoff.
Is it realistic to design a small battery-powered Bluetooth device with an app and manufacture 1,000 units for under $50,000?
It can be, if the scope is tight: a pre-certified Bluetooth module, one PCB, a simple app, an off-the-shelf or 3D-printed/low-cost-tooled enclosure, and certification for one market. Engineering alone for such a product typically starts around $25,000–$40,000 with an India-based team, and 1,000 units of inventory, tooling and certification can consume the rest quickly. Most founders either phase the product (launch with fewer features) or budget closer to $60,000–$100,000 for the full path to 1,000 units.
What is the difference between EVT, DVT and PVT?
EVT (engineering validation test) proves the electronics and firmware meet the functional specification, using boards close to final form. DVT (design validation test) proves the production-intent design, including the tooled enclosure, passes reliability and certification testing. PVT (production validation test) proves the factory can build it at yield, using the real production line, fixtures and process; PVT units are usually sellable.
How do I protect my idea and make sure I own the IP when I hire a design firm?
Sign a mutual NDA before sharing details, and make sure the development contract assigns all project-specific (foreground) IP to you on payment and grants you a perpetual licence to any pre-existing (background) IP embedded in the design. The contract should also list every deliverable you receive — native schematic and layout files, Gerbers, BOM with manufacturer part numbers, firmware source with build instructions, and mechanical CAD. If the idea is patentable, talk to a patent attorney about a provisional filing before any public disclosure or crowdfunding campaign.
Sources
- USPTO — Provisional application for patent — US provisional filing basics (checked September 2026)
- WIPO — Frequently asked questions: patents — novelty and prior public disclosure (checked September 2026)
- FCC — Equipment authorization — US radio-product approval routes (checked September 2026)
- European Commission — Radio Equipment Directive (RED) — EU conformity for radio products (checked September 2026)
Cost and duration ranges on this page are Rapid Circuitry planning estimates, stated with their assumptions above.