August 15th, 2007
The latest posting on the LaserMotive blog discusses ongoing problems they’ve had with their climber’s motor controller. These types of problems are typical engineering ones; I’ve seen many like it during my time of gainful employment. It’s what the engineering process is all about, and it’s why it will take real engineering skills to build something that has a chance in winning the climber/power-beaming competition (this year, or any year).
One engineering “skill” is employing testing, testing and testing some more. I remember when I was working at a company called InFlight Phone Corporation. Our product was phone and entertainment systems for commercial and private aircraft (my job was managing all of the software development). We were having a problem with the phone cord mechanism on many of our commercial plane installations - it kept failing. Our new corporate owner, MCI, brought in a “SWAT” team of people to help us make our product more robust. Some of their people were competent, some were not. One of the not-so-good ones got involved with the problem of the handset cord failures. He looked at our current solution and soon came up with a new one, one he was sure would solve the problem. He was so sure that he said it didn’t need testing - just get it out to the planes ASAP. Fortunately, cooler heads prevailed and, over his objections, we brought in an ME to do some real analysis and testing on it. The ME rigged up a mechanism that repeatedly pulled the cord in and out of its holder, simultaneously dragging it over a plastic arm, simulating the type of motion it would be subjected to on a plane. Within 12 hours, the new, “improved” design failed.
Testing, testing, testing - an absolute key ingredient to success…