In my last posting I advised designers to avoid relying on “rules of thumb” or computer coding when confronted with tough decisions. This admonition has its roots in one of the key moments in my own career as a gearbox designer. A sizeable team was assigned to an emergency upgrade of a product line that was not faring well in the marketplace. It was the company’s first foray into surface hardened gearing and the idea of using a ground carburized and hardened pinion with an unground induction hardened gear did not pan out. More gear grinders were on the boat from Europe and the revolving elements had to be improved and tested before they were put into service. Other internal changes, such as a shift from ball bearings to taper roller bearings, were also needed, so new “study layouts” and calculations were required. Marketing weighed in with specific horsepower capacity goals for each ratio of each size. There were twelve sizes with 36 ratio combinations in each size. Four of those units were my portion of the work load. Computer time was a limiting resource. I had thirty-six sizeable boxes of “punch cards” that I personally typed out as input to the main frame computer. Each night our “problem sets” would be messengered to the computer room at the main plant and run through the IBM 370 overnight. A single bent, damaged, or misplaced card would void the run and delay your report. Our messenger was notorious for dropping boxes of cards and stuffing them back in without telling anyone — so every designer was diligent about marking and securing their cards for reshuffling. Our team was having a terrible time meeting the goals marketing set for us. One day, however, a set I was struggling with miraculously healed itself and made the capacity target. I had retyped an input card to replace one damaged in a transport mishap; a careful check on it revealed that I had transposed a couple digits on the pinion outside diameter. This larger diameter, although violating the “old family recipe,” did not set off any warning messages and the strength capacity increased enough to meet the goal. My supervisor chewed me out for sloppy typing but did not specifically forbid me from investigating the beneficial effects further. I soon had all of my “problem ratios” meeting the marketing targets and proudly reported this at the team meeting. My reward was a private interrogation by the department supervisor and the gear software writers. It was an uncomfortable couple of hours, but it resulted in a revision to the old family recipe to include a “strength balancing routine.” An average designer respects boundaries. A great designer moves the boundaries. Aim for greatness.