The VW scandal

vw_small_whatWhat the heck? It was recently found out that VW sold lots of diesel cars that disabled emissions control elements when those cars weren’t being tested.   Practically every engineer I’ve known has owned a Volkswagen at some time in their lives, and it probably saddens us all that this has happened.

From a technical perspective, this probably wasn’t too difficult of a task.  One scenario – if the car was under test in a shop or on a testing stand, the steering wheel would not be moving, and the front wheels would be moving while the rear wheels were stopped (the ABS sensors would know this).   All easy enough conditions to detect in software, and turn on the emissions control elements full blast.  Likewise, if those conditions weren’t met, the car’s emissions hardware could be shut off or degraded.

Firstly, kudos to Dan Carder and his team at West Virginia University for doing the legwork on this, and discovering this horrific hack.   Those folks are a great example of how engineers can and should look out for the common good of the public.

Secondly, it wouldn’t surprise me one bit that someone in the C-suite (or near it) could sign off on this project.  There are plenty of examples in the world of executives going into a moral gray zone (or abandoning morals completely) to boost the bottom line; Enron is the classic case.  What does surprise me is that some engineer (or engineers), after listening to this idea, might have written the code that actually did this, with the active intent of defeating emissions regulations.

Now, it is very possible that this code was used in some sort of testing phase, so that before and after testing could have been done on a testing track, and somehow this code got into production vehicles.  That does sound plausible, on one level, but I would think that the software that goes into automobiles must meet a much higher standard than other ones that go into typical software.  For example, the software that goes into controlling the ABS braking system would probably be rigorously tested and debugged.  However, mistakes do happen, and this might, just *might* be the reason that this deactivation happened.   It will be up to VW to show that this was the case.   For certain, the lawyers will have a field day with this issue.

Software has become ingrained in our modern lives, and as the control of equipment and real world devices by computers becomes even more ubiquitous, it will be up to the people who program those computers and embedded systems to have a good moral compass, and do the right thing.

It may be that VW’s reputation is ruined for a long time.  Morton Thiokol and the space shuttle solid rocket boosters, Ford and the Pinto, GM and those ignition switches; these kinds of gross lapses in judgement are not forgotten easily, by the public, and by engineers especially.


This entry was posted in articles. Bookmark the permalink.