In the four OVA Solutions proposals that priced the whole path from an idea or early prototype to a production-ready design or pilot batch (three medical devices and one consumer product, 2025 and 2026), the estimate was 4,000 to 6,000 engineering hours; the costliest route in one of them reached about 7,200. Taking a concept or proof of concept to a few working prototypes was estimated at 1,000 to 2,500 hours. In the four Class II and sterile proposals that broke it out, documentation and verification work was estimated at 15% to 31% of the hours. And of the roughly 20 founders who named a budget on a call, about half named one below what the scope they described would need, some by a factor of 3 to 10.
Hours by starting point and goal
Each range is the engineering hours written in OVA proposals for that kind of request: electronics, firmware, mechanics, testing, documentation and project management. Materials, test labs and regulatory fees are not in these numbers.
| Starting point and goal | Proposals | Estimated engineering hours |
|---|---|---|
| Idea or proof of concept to a demo or investor prototype (1 to 5 units, mostly off-the-shelf or printed parts) | 6 | 80 to 700; about 450 to 700 for the five medical and research demos; the low end is a simple consumer gadget |
| One feasibility gate on the riskiest part (when one physics question decides the device) | 6 | 400 to 1,000; typically 400 to 560 |
| Concept or proof of concept to 3 to 10 working prototypes, custom electronics, no DFM | 8 | 1,000 to 2,500 |
| Device on the market to a redesign (obsolete parts, new features) | 1 | 1,200 to 3,000 |
| Existing prototype to pre-production units with a design history file | 2 | 2,100 to 3,240 |
| Proof of concept to a production-ready or submission-ready design | 2 | 3,000 to 4,000 |
| Idea or early prototype to a production-ready design or a pilot batch | 4 | 4,000 to 6,000; one costlier route up to about 7,200 |
| Large systems such as ventilators or glucose monitors | published breakdown | 10,000 to 30,000 |
Where the hours go
Almost every proposal was split into phases with a gate between them. Most proposals for a full program opened with discovery as a separate step (about 14 of the 35 had one); small demos and single feasibility gates usually did not. This is the recurring structure:
| Phase | Typical hours | Share of a full program |
|---|---|---|
| Discovery: requirements, architecture, risk list, plan for the next phase | 100 to 336 (median about 180) | 5% to 15% |
| Feasibility gate, only when one physics risk decides the device | 400 to 1,000 (typically 400 to 560) | 10% to 30% when present |
| Design and first prototype: electronics, firmware, mechanics, first build | 700 to 1,600 | 35% to 50% |
| Next prototype round: target form factor, fixes, bench and on-body tests | 200 to 1,000 per round | 15% to 25% |
| Documentation and verification preparation (medical programs) | 400 to 1,250 | 15% to 31% of a Class II or sterile program |
| Design for manufacturing, transfer and pilot build | 1,000 to 2,300 | adds about 30% to 60% on top of the prototype work (2 proposals that priced it separately) |
| Project management and coordination | line item or spread over phases | 6% to 18%, usually 10% to 15% |
Discovery was usually planned at 3 to 8 weeks. Its output is a requirements document, an architecture, a list of risks and an estimate for the next phase, so the first estimate based on a review of the design comes after it.
The design-for-manufacturing hours above come from three proposals. Lisa Voronkova’s published breakdown puts design for manufacturing for an average product at 2,500 to 4,000 hours.
Medical vs wellness
The biggest difference in hours between a medical device and a wellness product in these proposals is the documentation workstream that runs from discovery onward: design controls, risk management under ISO 14971, software documentation under IEC 62304, preparation for IEC 60601 testing and verification protocols.
- In the four Class II and sterile proposals that broke this work out, it was 15% to 31% of the estimated hours: about 15% on a Class II wearable, 19% to 26% on a powered Class II system with IEC 62304 Class B software, 20% to 29% on a Class II accessory set with a full design history file, and about 31% on a sterile single-use device. The lower figure in each pair counts the documentation phase only and the higher one counts all QA and regulatory hours; the sterile figure also includes test-method development and bench verification.
- Wellness demonstrators carried no design controls or regulatory documentation. The one wellness product redesign that included compliance preparation (safety, EMC and CE readiness) put about 8% of its hours there.
- As a rough guide from these proposals: documentation and verification add about 20% to 30% to a program’s hours. Industrialization under design controls (verification, design for manufacturing and a pilot build) is a separate step; in one proposal it added another 36% to 57% on top of the prototype work.
On the sampled calls, 66% of the prospects described a medical device and 27% a wellness or consumer health product. Wellness founders arrived earlier (10 of 19 had only an idea, against 10 of 46 medical founders) and wanted discovery or a feasibility step first about three times as often (10 of 19 against 8 of 46). At least 7 planned to launch as a wellness product first and seek FDA clearance later.
What changes the number
- A cellular radio on a battery was the largest unknown in the one proposal that had it and got its own phase of about 500 hours. Bluetooth was routine.
- Each additional sensing modality widened verification and regulatory scope; one program’s all-in-one route took 14% to 40% more hours than a phased route.
- Custom compact electronics instead of modules cost about 50% more hours and three more months on one data logger.
- Each extra board spin or prototype round cost 200 to 500 hours in most proposals, and up to about 1,000 on complex devices.
- Mechanical risk, such as a custom pump instead of an off-the-shelf one, was named in one proposal as the item that could swing budget and timeline the most.
- Units to build: 2 to 3 hours per printed device, about 20 to 25 hours per electromechanical unit.
- Compressing the schedule with parallel workstreams cost 6% to 26% more hours for two months less.
Why estimates move
The number given on a first call is not the number in the proposal. On a sterile single-use device the estimate went from about 2,000 hours on the call to 3,000 hours to design freeze, plus 1,000 hours for a pilot batch, after the engineering lead reviewed the CAD. On a consumer device a founder’s own guess of 800 hours became about 2,400 hours in the first proposal and about 5,760 hours in the fourth version, a fixed-price 12-month plan written after the scope, timeline and payment terms had changed.
That is why proposals carried a margin: a tolerance of plus or minus 15% until discovery in some, a 20% risk buffer in another, or a not-to-exceed ceiling about 17% to 33% above the estimate. One grant budget suggested 20% to 30% contingency.
What founders bring and ask
| In the call sample, founders arrived with | Share of 70 |
|---|---|
| Only an idea: a deck, a sketch, a render | 30% |
| A proof of concept: a bench or development-kit setup, often academic | 24% |
| An existing prototype that needs the next iteration, units or production | 36% |
| A device on the market that needs a redesign | 10% |
What they wanted at the end: working prototypes (61%), design for manufacturing and production transfer (31%), units for a clinical study or pilot (27%), discovery or feasibility first (26%), a redesign or next generation (20%), regulatory documentation (16%, all of them medical).
Budgets: only 10 of the 69 device prospects who were asked about budget on the booking form gave an amount; about 20 of the 70 named a figure on the call. Roughly half of those figures were below what the scope they described would need, some by a factor of 3 to 10. Eight founders said they wanted the number in order to know how much to raise.
The questions asked most often on the sampled calls, most frequent first:
- How much will it cost and how long will it take, so I know what to raise?
- What is your rate, and is the price hourly or fixed?
- Can we pay later, in equity or with a revenue share?
- Do you manufacture, and how many units can you build for a trial?
- Do you handle the FDA side: design history file, 510(k), ISO 13485?
- Have you built something like this before?
- Should we build from scratch or start from something off the shelf?
- Where is your team, and is it safe to work with engineers in Ukraine?
- Who owns the IP, and do we sign an NDA first?
Our answers to the two most practical ones: we prepare the engineering side of a submission (design history file, risk management inputs, verification reports), and the regulatory strategy and the submission stay with the client or its regulatory consultant. We build prototypes, golden samples and batches of up to 200 units; volume production goes to a contract manufacturer, and we support the transfer.
What this means if you are planning a device program
- Budget the whole path in hours, not only the first prototype. A working prototype is only part of the path: in the proposals that priced both, design for manufacturing, verification and documentation added another third to roughly as many hours again.
- Buy discovery first. It ends with requirements, a risk list and an estimate for the next phase that rests on a design review, not on a first call.
- Write the documentation while you design. In the Class II and sterile proposals it was 15% to 31% of the hours. Our experience is that written at the end, it costs more.
- Prototype hardware is the small line. In one proposal, parts ran about $1,000 to $1,600 per wearable unit; in another, $6,000 to $9,000 for a single RF demonstrator. Prototype materials typically ran about 10% of the engineering cost and were billed at cost. Tooling, sterilization and test labs are a separate and much larger budget once a device goes to a pilot batch.
- Raise for a result you can show: working prototypes, units for a clinical study, or a design your manufacturer can build.
To turn these hours into a budget for your own device, use the medical device development cost calculator. For what the rate covers and what is paid to others, see how much medical device development costs.
Method and limits
- Proposals: 35 written estimates that OVA Solutions and its group companies sent to prospects and clients from July 2024 to October 2026, found by searching sent mail and shared files for estimates with hours; it is not every estimate OVA sent. 34 state hours; in a few, hours are derived from the quoted price. About 20 are medical devices or medical research instruments and about 15 are wellness, consumer or lab products. Most were not signed, and acceptance was not checked for all. They span wearables, orthopedic and rehabilitation devices, neurotechnology, diagnostics and surgical equipment.
- Calls: 70 prospect calls from February 2025 to October 2026, most of them first calls (about 11 were second calls or returning contacts). They were chosen from about 500 recorded call notes as a sample stratified by the development stage founders ticked when booking, so the shares of 70 describe this sample, not all OVA prospects. Starting point, goal and medical or wellness status were coded by OVA from the notes. Budget answers come from the booking form (69 device prospects were asked) and from the calls.
- Anonymized: no client names, no device that could identify a client, and no engineering prices or rates from individual contracts; the only dollar figures are prototype material costs, which are billed at cost. We report hours, because hours compare across firms and contracts; multiply by any rate to get a budget.
- Limits: these are estimates, not hours spent, and were not compared with actual hours; several rows rest on one to four proposals; the mix reflects OVA’s clients; ranges are what was written, not a statistical model. The analysis was done by OVA and was not independently audited, and OVA sells the discovery and engineering work described here.
Download the aggregated tables: CSV, ova-medical-device-development-hours-2026.csv. Free to use with attribution: OVA Solutions, “Medical device development in hours”, October 2026, with a link to this page.
Questions
How many engineering hours does a medical device take?
In the four OVA proposals that priced the whole path from an idea or early prototype to a production-ready design, the estimate was 4,000 to 6,000 hours, with one costlier route near 7,200. Taking a concept or proof of concept to a few working prototypes was estimated at 1,000 to 2,500 hours in eight proposals. In Lisa Voronkova’s published breakdown, large systems such as ventilators or glucose monitors take 10,000 to 30,000.
How much of a medical device program is documentation?
In four Class II and sterile proposals, documentation and verification work was 15% to 31% of the estimated engineering hours. Wellness demonstrators carried none; one wellness product redesign carried about 8% for compliance preparation.
How long is a discovery phase?
Usually 3 to 8 weeks and 100 to 336 engineering hours, with a median of about 180. It ends with requirements, an architecture, a risk list and an estimate for the next phase.
Why does the estimate change after the first call?
Because engineers see the design. In one case the call estimate of about 2,000 hours became 3,000 hours to design freeze plus 1,000 hours for a pilot batch after a CAD review. Before discovery, OVA proposals carried a margin for this: a tolerance of plus or minus 15%, a 20% risk buffer or a not-to-exceed ceiling.
Can I use these numbers in an article or a pitch deck?
Yes, with attribution to OVA Solutions and a link to this page. The aggregated tables are available as a CSV file.
Related pages
Want these numbers for your device?
Tell us where your device is today. Lisa Voronkova will go through it with you; after an NDA, our CTO reviews the design and we send an estimate in hours by phase.
By Lisa Voronkova, CEO of OVA Solutions. Published October 10, 2026.