The Back Porch
The Back Porch
To Skin a Cat
By
Johnathan W. Carson
Originally published July, 2026
Many of the best engineering lessons I ever learned didn't come from my own ideas. They came from my mentors. In this case, it was Doug Packard, who was probably influenced by one of his own mentors.
Early in my career, whenever I came up with a design that looked promising, I'd start detailing it. After all, I had a solution. Why waste time looking for another?
Years later, I learned Packard had a simple rule: before detailing anything, develop at least three concepts for every design problem. I don’t remember exactly how Doug phrased this particular rule, but I adopted it to mean three fundamentally different approaches, not just a variation on a theme. Suppose you need a mechanism that provides linear motion. Your first idea might be a brushless motor driving a ball screw. Replacing the brushless motor with a stepper motor, or the ball screw with an Acme screw, does not give you a second concept. Those are variations of the same basic approach. A variation is not a new concept. A genuinely different second concept might use a voice coil. A third might use an inchworm actuator. Now you are exploring three different ways of producing the motion, not merely changing components within one architecture.
I’ll admit, it can be hard sometimes to come up with three different approaches, but take some time and do your best. Maybe you can't always do it. Honestly, sometimes I struggle too. This was especially difficult early in my career. It is hard to imagine three fundamentally different mechanisms when you have only been exposed to a handful of ways to solve various problems. This becomes much easier as your design toolbox fills with experience. That is one reason mentors, coworkers, old design reports, textbooks, and even hardware catalogs can be so valuable. Sometimes creativity begins by learning what possibilities already exist. Don't hesitate to borrow good ideas or adapt old solutions in new ways. This may take some time, but it is time very well spent. It will pay you dividends in the future.
The reason I advocate this discipline is that it forces you to spend time thinking deeply about the problem, and often from different vantage points. After you've done this several times, you’ll start to see that your first idea is not necessarily your best. Sometimes it is that second, or even third idea that has the most merit. Usually the best solution is a concept that combines the strongest features of two or three earlier ideas. You will never discover that if you just stop developing concepts the moment you come up with a single idea and start detailing.
Schedule pressure can make this multi-concept approach difficult. Once a customer or manager hears that you have a concept that appears workable, the natural reaction may be, “Good. Start detailing it.” From their perspective, it’s understandable. From the engineer’s perspective, however, a little more exploration may prevent weeks or months of rework. For that reason, I am careful about presenting any particular concept as the solution before I have finished examining the alternatives. I may say that I'm evaluating several promising ideas rather than declaring the first feasible one the winner.
One of the clearest examples of this came late in my career, when I was designing the Focus Control Mechanism for the Roman Space Telescope. The mechanism required approximately ±1 millimeter of travel with a resolution of five nanometers. That meant resolving motion roughly 400,000 times smaller than the total travel—a surprisingly difficult combination to achieve. Those requirements created an extraordinary dynamic range, and none of the first concepts felt entirely satisfactory. By that point in my career, developing multiple concepts had become second nature. I didn't think twice about continuing to produce new ideas. In all, I developed six or seven different approaches before selecting the one I believed offered the best overall solution. At the Preliminary Design Review, I presented the baseline design along with three of the more promising concepts that had led to it. Their purpose was not to solicit a choice from the review board, but to show the design space that had been explored and why the selected approach represented the best path forward. Even then, the final mechanism was not identical to any one of those concepts. It combined elements from two different approaches. It was relatively simple, but reaching that simplicity required exploring a much larger part of the design space first.
I now realize that Packard's rule wasn’t really about generating three concepts. It was about resisting the temptation to fall in love with your first idea. Once you've spent weeks detailing a concept, it becomes your concept. Confirmation bias quietly begins to creep in. Every problem gets rationalized away because changing directions now feels like throwing away all that work. The further down that road you go, the harder it becomes to turn back.
Developing three concepts early on is simply the discipline that prevents that trap. It forces you to compare alternatives before you become emotionally invested in any one of them. The best designs rarely emerge fully formed. They emerge from comparing ideas, borrowing strengths from each, and giving yourself permission to abandon yesterday's favorite in favor of something better. If you can learn to do that consistently, you'll usually end up with a simpler, more elegant, and often more robust solution.
Takeaways
Your first workable idea is rarely your only good idea.
Develop fundamentally different concepts before committing to one.
The best design often emerges by combining the strongest ideas from several concepts.
Explore Further
Packard’s Laws (coming soon)
Related Conversations
You Brought Your Machete, Right?