Classification Sets the Order of Magnitude

Before you can estimate a schedule you need to know which regulatory pathway the device takes, because the difference between them is years, not months.

  • Class I, exempt. Nine to eighteen months from concept to market. No premarket submission, but establishment registration, device listing, labeling, and quality system obligations still apply.
  • Class II with a 510(k). Two to three and a half years is typical for a device developed from scratch. Roughly eighteen to thirty months of development and testing, then four to eight months of submission and review.
  • Class II with no predicate, via De Novo. Three to four and a half years. The review clock alone commonly runs ten to fifteen months in practice.
  • Class III with a PMA. Five to eight years and frequently longer, dominated by the clinical study rather than the engineering.

Getting the class right is the first schedule decision you make, and getting it wrong resets the plan. Start with FDA medical device classes, and if a similar cleared device already exists, identify a candidate predicate in the same week rather than the same quarter.

Feasibility and Concept: Three to Six Months

The goal of this phase is to prove the core mechanism works and to lock the regulatory strategy. Deliverables are a documented user need, an intended use statement written in the language the submission will use, a proposed classification with product code and pathway, a preliminary hazard analysis, and a bench prototype that demonstrates the physical or biological principle.

Three months is achievable when the technology is understood and a clear predicate exists. Six or more is normal when the mechanism itself is novel. The one thing that must not be deferred is the intended use statement: it determines classification, which determines testing, which determines cost and calendar.

Design Controls and Risk Management Run the Whole Way

From the moment feasibility ends, the program runs under design controls: design inputs, outputs, reviews, verification, validation, and change control, all traceable. Alongside them runs risk management, which is not a document you write once but a live file updated at every design change and every test failure.

Teams that set this up at the start add perhaps 10 to 15 percent to engineering effort. Teams that reconstruct it afterwards spend three to nine months rebuilding evidence for decisions nobody recorded, and that reconstruction is the single most common cause of a first device shipping a year late. The mechanics are in FDA design controls and ISO 14971 risk management.

Detailed Design and Verification: Nine to Eighteen Months

This is the bulk of the engineering: mechanical design, electronics, embedded software, industrial design, and human factors, converging on a design freeze followed by verification testing against every input.

The long-lead items inside this phase are worth naming because they do not compress:

  • Biocompatibility. Twelve to twenty-six weeks for a typical panel on surface-contact devices, longer for implants, and the samples must be made from final materials with final processing, including sterilization. Test selection is covered in ISO 10993 biocompatibility testing.
  • Electrical safety and EMC. Eight to sixteen weeks for a full IEC 60601 program including one retest round, which most first devices need.
  • Sterilization validation. Twelve to twenty weeks, plus accelerated aging for shelf life, where a claimed three-year shelf life takes several months of accelerated aging even under ASTM F1980.
  • Summative usability study. Two to four months including recruiting, sessions, and analysis, run on production-equivalent units.
  • Software. If the device contains software, documentation level drives the evidence burden, and connected devices add a cybersecurity package.

These run in parallel, but only if the design is frozen. Every design change after samples go to the lab restarts the affected test.

Clinical Evidence and the Submission

Most 510(k) devices need no clinical data; substantial equivalence is argued from bench and animal testing plus the predicate comparison. When clinical data is required, add twelve to thirty months for protocol, IRB approvals, enrollment, follow-up, and analysis, and an IDE application if the study is significant risk. That decision is examined in clinical trials for medical devices.

For the submission itself, plan six to twelve weeks to assemble a 510(k) and four to eight months of elapsed review time. The statutory clock is 90 FDA days, but the clock stops when the agency issues an Additional Information request, which happens on a large share of submissions and typically costs two to four months. The steps are in the FDA 510(k) submission process, and the no-predicate route is described in the De Novo pathway.

Where Timelines Actually Slip

Late classification decisions that change the test plan after money is spent. Design changes after test samples ship, which restart biocompatibility or EMC. Underestimating the quality system, which has to be operating before design transfer, not after. Sequential rather than parallel testing, which can add six months for no technical reason. Running out of money mid-program, the most common failure of all, which is why the budget in medical device development cost should be raised against the pessimistic schedule, not the optimistic one.

Build a Schedule You Can Defend to Investors

Projects House plans medical device programs with the regulatory pathway, test lead times, and design controls mapped onto one dated schedule, so you know which items are on the critical path before you commit. Send your intended use and current stage through our contact form.