When it comes to meeting a customer’s sensing requirements, there is typically more than one way to solve the problem, and there is also more than one technique to get from problem to solution. For engineers, who can turn ideas into reality, the techniques used may be subconscious, like the cook that bakes a cake without a recipe or measuring. But let’s be a little introspective here as to what techniques are used in the product development process.
These steps are commonly included in product development:
• Idea generation – Brainstorm
• Idea screening – Sift
• Development – Build, evaluate, refine
• Validation tests
• Product launch
Evaluation or testing methods are needed in several of these product development steps, specifically in screening ideas, developing ideas into a physical product and validating that the product meets requirements. For me, evaluation methods are one of the key pieces of product development. Without good evaluation methods, a sub-optimal idea may be chosen for development and the development may be sub-optimal. The product could also not meet customer requirements, or a product is launched that everyone thinks meets the customer needs but does not.
Traditional evaluation methods involve building parts and measuring their activation and deactivation switch points using mechanical fixtures and micrometers or calipers. I call this bench testing. This is useful and valuable in all three above-mentioned product development steps where evaluation is needed.
Another evaluation method is using computer modeling to predict product performance before a part is built. This is not particularly useful for validation testing, but it can be a great help in idea screening and in developing and refining the product. In addition, as computer technology continues to advance, the ability to “number crunch” more and more data allows more thorough evaluation of more design variations.
However, computer modeling does not replace bench testing. They are both needed and complement each other, but sometimes one method is better than the other to get from point A to point B.
The use of computer modeling requires a somewhat different skill set and mindset compared to bench testing. Instead of “hands on” work, more abstract thinking and a computer skill set is required. To automate repetitive modeling tasks, programming skills are valuable. Part of the mindset change is trusting computer results to accurately predict the behavior of real parts. Building trust involves verifying modeling results match real part performance. This must be done over and over until a history of success is developed.
Computer modeling can, and should, produce both quantitative and qualitative results. By quantitative, I mean numeric results that can be used to evaluate different alternatives, optimize a design and verify that a design meets the requirements. By qualitative, I mean visual results that can provide insight into the how a design operates. Magnetic fields are complicated. They curve, they vary in intensity, they vary in direction and they are invisible. Visualization of magnetic fields through qualitative results helps to understand why a design works or doesn’t work, and how to improve it.
As technology advances, some things become less attractive and other things become more feasible. Success requires taking advantage of new technologies, new methods and new ideas. The only constant is change.
