1. Three development paths, not two
Programs often compare “off-the-shelf” and “custom,” but a modified platform is a useful middle path. Existing optical architecture may be retained while mount, filter, coating, barrel, focus or sealing is adapted.
Classify the requested change before estimating timing and cost. A small external change is different from changing FOV, image circle or optical performance.
- Existing qualified platform
- Platform with controlled modification
- New optical and mechanical design
- Camera-level co-development
2. Measure the real fit gap
Create a requirement matrix covering sensor, FOV, MTF, distortion, spectrum, package, environment, assembly and validation. Mark each item as met, adjustable, uncertain or not met.
The number of gaps matters less than their severity. One image-circle or thermal-focus conflict may force a new design, while several interface changes may be manageable.
- Optical and sensor fit
- Mechanical and sealing fit
- Environmental and validation fit
- Manufacturing-process fit
3. Compare evidence, not only development time
A mature platform may have tolerance data, process learning, reliability history and production controls that reduce program risk. A new design can optimize the exact target but must build that evidence through prototypes, tooling and validation.
Ask which evidence transfers unchanged, which evidence is partially reusable and which tests must be repeated after modification.
- Tolerance and production history
- Reliability evidence
- Tooling and process readiness
- Required revalidation scope
4. Model schedule and commercial risk honestly
Development time includes requirement freeze, optical design, mechanical design, prototype builds, tooling, test loops, customer validation and production ramp. Late sensor or package changes can reset multiple stages.
Volume, lifetime, target cost and forecast confidence influence whether custom tooling and validation investment is justified. Use scenario ranges rather than a single optimistic date.
- Requirement-freeze confidence
- Prototype and tooling loops
- Validation and customer approval
- Volume, lifetime and investment case
5. Use a gated decision process
Begin with a documented platform screen. If gaps remain, run a short feasibility study before committing to full custom development. Freeze interfaces and success criteria at each gate.
The output should be a clear recommendation, open-risk list, responsibility matrix and next validation step—not simply a lens part number.
- Gate 1: requirement and platform screen
- Gate 2: feasibility and risk closure
- Gate 3: design / tooling commitment
- Gate 4: validation and SOP release