When it comes to product development, it is easy to get stuck in a prototyping loop. Ironically, driving the prototyping process too fast is the usual cause of the trap. Trying to optimize every design variable simultaneously means that the overall project flounders. In optical engineering, a prototype trap can be created by setting tight performance specifications too early.
Below is advice and best practices for the three stages of prototyping: conceptual design, preliminary prototyping, and final prototyping.
1. Conceptual Design
Conceptual design in optics entails the early modeling and optimization of the instrument. There is rarely a “best” solution, but instead a range of possible solutions. At this stage, it is advisable to pursue several concepts. Keep the design and performance specifications loose. Loose specifications for concept parts means the parts will cost less, and you won't blow your budget early on. It also enables you to test out more concepts. Another advantage of loose specifications is that you can start to get a feel for which design parameters most affect final performance.
2. Preliminary Prototyping
Preliminary prototyping involves adjusting single design variables in a systematic, iterative approach. Preliminary prototyping should always include two or three different, feasible conceptual designs. Performing "preliminary prototyping" on only one design concept is actually jumping to final prototyping based only on conceptual measurements. Frequently that leap fails, and you find yourself in a loop of design, prototype, fail, repeat. If, instead, there are multiple workable concepts, you can move forward from a failure, rather than looping back.
A common error during preliminary prototyping is attempting to optimize one variable at a time with the notion that moving aggressively will lead to fewer prototypes—quite the opposite. You can end up landing in an optimization loop, where perfecting the performance via one variable undoes the optimization of the preceding variables. The result is an increasing number of unworkable prototypes. Instead, adjust each variable by a small amount to determine how each design variable affects overall system performance. From there you can progress each variable toward optimization without over-shooting the optimum configuration of variables. This sequential process can seem tedious and expensive, but by pursuing this approach, the design moves steadily forward.
At the end of preliminary prototyping you will have one or more workable concepts, or you will have determined that none of the design approaches work and need to be re-specified. If more than one design has survived, a single design concept must be chosen to move on to final prototyping. Pushing multiple designs into "final" prototyping means that you are actually still in preliminary testing, but you have elected to tighten the specifications beyond the performance requirements. That is a perfectionism trap. You do not need the best design to move into final prototyping; you need a design that is good enough.
3. Final Prototyping
Final prototyping is the refinement of manufacturing methods until the product can be fabricated in the quantity, quality, and price required. No design should ever reach final prototyping until it has been proven, through preliminary prototyping, that the design performs within specification. Final prototyping is not a test. A product should never “fail” final prototyping. Looping from final prototyping back to any earlier stage is incredibly expensive and often fatal to a project.
Developing component-level performance specifications should be a systematic step-wise process that moves steadily toward peak performance, rather than a circular process that loops around peak performance. The best way to minimize looping is to communicate with your optics manufacturer throughout the process. By understanding the intent of each prototype, the manufacturer can move your project forward while avoiding set-backs.
To ensure a smooth prototyping process, RPO offers a variety of prototyping capabilities for custom optics. Our in-house expertise includes techniques for rapid protomolding that can cut lead time in half, single-point diamond turning of chalcogenide glasses for IR optics, and optical assembly prototyping for more efficient scale-up in optics.