GM Names Names, Suspends Two Engineers Over Ignition-Switch Safety 236
cartechboy (2660665) writes "GM said it has placed two engineers on paid leave in connection with its massive recall probe of 2 million vehicles. Now, GM is asking NASA to advise on whether those cars are safe to drive even with the ignition key alone. Significantly, individual engineers now have their names in print and face a raft of inquiries what they did or didn't know, did or didn't do, and when. A vulnerability for GM: One engineer may have tried to re-engineer the faulty ignition switch without changing the part number—an unheard-of practice in the industry. Is it a good thing that people who engineer for a living can now get their names on national news for parts designed 10 years ago? The next time your mail goes down, should we know the name of the guy whose code flaw may have caused that?"
Comparison to code bugs a bit flawed (Score:5, Informative)
The fine article submission asks:
One key difference here is that the engineer(s) responsible for redesigning the switch and not changing the part number were not just implementing an everyday change that happened to be buggy. By not changing the part number, their actions are more akin to trying to fix a known bug that has exposed the company to huge potential liabilities, and then hacking the version control system to make it look like the bug was never there, in full intentional pursuit of obfuscation and ass-covering.
Cheers,
changing part without changing number is common (Score:5, Informative)
I'm not saying that it's good, and this case is an example of exactly why it can be a bad idea to do this, but changing part numbers has a lot of overhead (inventory management of multiple part numbers, all the manuals that now refer to the wrong number, etc)
If it's expected that the new part is significantly different than the old one, then it's worth all the pain, but if it's not expected to be significantly different (just cheaper to build, or more reliable when nobody expects series reliability problems with the old one, etc) its not completely insane to just change the design and keep the same part number.
If you want to be really paranoid, you track each batch of parts produced as a separate item, because minor things like the temprature that day could theoretically affect something. In medical and aerospace industries, this sort of tracking is done (which is one of the reasons why 'simple' things are so expensive in those industries)
but in the automotive industry that level of tracking is just not done, and it's very common for parts to be substatuted with no notice.
In the computer industry, it's unfortuantly common for some manufacturers to make what many people consider major changes (like changing chipsets) without changing the part number.
David Lang
Professional responsibility (Score:5, Informative)
Engineers are professionally certified with professional responsibility, if they aren't doing their job it's criminal and names need to be named. Just as a physician working for a hospital is named for accusations of negligence.
It's not obvious if that's relevant here, but if someone tried to pass themselves off as a professional engineer and aren't that's a problem, if someone who is a professional engineering violated the ethical principles that's a problem too.
Re:Hero ? (Score:4, Informative)
That said, naming names of an engineer is a really bad precedent. What is the goal GM is trying to achieve here. Do they want people to go break the guy's windows? Burn down his house? Call him in the middle of the night or deliver pizza? Apart from potentially removing the guy's livelihood for the remainder of his life because no-one wants to hire 'that guy' ever again, and a lot of abuse being targeted his way, what will this achieve?
Why exactly is it a bad precedent?
The names of everyone involved are going to come out anyways, with all the possible consequences you described.
Our judicial system is usually exceedingly unwilling to pierce the corporate veil and directly hold bad actors responsible for their choices.
So I'm perfectly happy with a society that aggressively shuns those people, regardless of judicial outcomes.
I'm *guessing* GM's goal is to scapegoat a few responsible parties as early as possible,
so that when the management failures are unmasked, there won't be as much heat and vitriol.
Re:Hero ? (Score:5, Informative)
Reuters [reuters.com] has a longer story that explains the background. Digrigio testified in the Senate that he did not know of the issue. Later senate dug up documents implying the opposite.Altman did something similar (but not nearly as bad) in front of a Jury.
Re:changing part without changing number is common (Score:5, Informative)
Since I work for an automotive OEM.... When this is done, there is an Engineering Change Order documenting the change and why it was implemented. We don't change anything without first getting the approval of the customer; and, invariably they will want all the relevant DV and PV testing redone. Huge effort and pain. All of this is well documented and nothing ships until we have final approval from the customer.
The part number may not change; but, the part revision level will. PN 123456 RevA will become PN 123456 RevB. We treat it as the "same" part number but will only ship the latest revision once we have customer approval. As for tracking, I don't know how our customers tracks the change internally; but, I can tell you which batch, serial number, and date code the new revision started shipping.
NOT engineer's fault (Score:2, Informative)
Everyone is SO fscking focused on the engineers. Yes, I am one. Here's my view:
What about part inspection and quality control? That switch has a detent holding force spec. Does NOBODY inspect batches of switches? No random sample tests? Doesn't someone do a final check on the car? Someone sits in it and turns the key, right? Didn't anyone notice the detent force was lacking?
If the switch detent holding force is this critical, then _EVERY_ switch should have been tested. It would be trivial to make a test jig and would be a 2-second automated test.
There are _many_ reasons this could have happened, accidentally or intentionally. It's not only engineers who can make changes. There are many examples in history where the maker / builder deviates from the engineered plans. St. Louis skywalk comes to mind: https://en.wikipedia.org/wiki/Hyatt_Regency_walkway_collapse [wikipedia.org]
Re:Hero ? (Score:5, Informative)
Actually, the ethical course is to blow the whistle. Sure you will loose your job and probably never work again, but .... Yeah, there are no up-sides to this one...
Re:changing part without changing number is common (Score:3, Informative)
The document linked in the article shows several revisions to the switch assembly without changing GM p/n (revs A, B, C1, C2, C3 and D).