The General Motors ignition switch fiasco remains in the headlines, having cost many lives, millions of dollars, and a number a careers. What seems to engineers as a relatively simple issue becomes very complicated in light of company politics, legal concerns, and accident settlements. As one colleague tersely put it: “Just tell people to stop hanging four pounds of junk on their key fob and the switches are just fine.”
The engineer in me wants to endorse this approach, but unfortunately the general public doesn’t see it so simply. We have a responsibility to anticipate misuse of our products and to do what we can to prevent misuse from harming people. This puts the engineer at odds with his employer’s need to limit liability and keep failures quiet.
We see this dynamic in action at AGMA technical committee meetings. A problem will come up in conversation and people will immediately start backpedaling on how they happen to be so well informed on it. It is an interesting bit of maneuvering — necessary for legal purposes but detrimental to solving the problem.
Closer to home, there are a number of ongoing issues bedeviling gear makers these days. One of them is grinding burn; despite claims to the contrary, gears occasionally get burned during tooth grinding. Instead of making a combined effort to identify the conditions that cause it and to develop methods of minimizing it, we pretend it only happens to “the other guy.”
A colleague has been trying to raise money for a test program on grinding burn, with an eye toward developing rework methods suitable for salvaging very expensive parts. Unfortunately he is not getting much support, because it is “the other guy’s” problem. Many potential funders are quite interested, however, in seeing his rework methods if and when they are developed. Without disclosure and evidence sharing, it is more difficult to solve problems.
Perhaps we need a South African-style “Truth and Reconciliation Commission,” where companies can get indemnification for previously undisclosed technical failures.