Report: Multiple Versions of Cheating Software Developed by Volkswagen

Aaron Cole
by Aaron Cole

Reuters reported that sources within Volkswagen and its ongoing investigation have said the automaker created multiple versions of its “defeat device” to cheat emissions tests.

The news agency said a manager within Volkswagen and an official close to the external investigation ordered by the automaker have revealed the multiple programs, which were developed for four different engine types.

If true, the multiple emissions programs could indicate a widespread cheating program — stretching nearly a decade — that could have needed funding to continue, which would be in stark contrast to the “rogue engineer” explanation offered by executives so far.

According to the report, diesel emissions experts said the automaker would have had to adapt its software for multiple versions of its engines, and for lean-trap nitrogen oxides filters (LNTs) and selective catalytic reduction systems.

Volkswagen of America chief Michael Horn admitted in his Oct. 8 testimony that cheat devices in the U.S. weren’t identical to the ones used in Europe, but didn’t elaborate.

“Since the standards are different, my understanding is that the defeat devices in those (European) cars are as well,” Horn said, according to Reuters.

Multiple versions and generations of software could potentially be damning for the company. If executives are implicated in the plot to cheat emissions tests, fines from governments could be significantly higher.

“The more higher-ups that are involved, the more the company is considered blameworthy and deserving of more serious punishment,” Brandon Garrett, a corporate crime expert at the University of Virginia, told Reuters.


Aaron Cole
Aaron Cole

More by Aaron Cole

Comments
Join the conversation
10 of 53 comments
  • NickS NickS on Oct 19, 2015

    It would most definitely have to be several different algorithms or implementations. You have different chassis, different engine families and codes, different exhaust treatment systems, different sensors, ... At a bare minimum, some at least minimal hand-tweaking would be needed for even minimal differences across vehicles, and that would trigger some subset of testing cycles to be run to verify the changes. If the differences go beyond a certain point, the original high-level design of the cheat-mode algorithm will need to be revisited, and re-architected, and that will certainly require re-development. There will be budgets and resources attached to each of these efforts. The Reuters piece shows a lack of understanding of software engineering, or how control systems work. I don't how many of these folks have ever come close to a software developer, but even if they have written a few thousands of Lines Of Code, they could not appreciate the complexity of a system of hundreds of millions of LOC.

    • See 1 previous
    • WildcatMatt WildcatMatt on Nov 04, 2015

      @callmeishmael Unless they use TFS, in which case all bets are off.

  • Callmeishmael Callmeishmael on Oct 19, 2015

    When this story broke I wondered how VW slipped the diesels past on-road testing. Simple: on-road testing doesn't seem to be a part of anyone's regulatory scheme. What the...? That's a hole you could drive a bus, or several million VW diesels, through. This is what happens when technically illiterate legislators cook up laws regulating things they know little or nothing about.

    • See 1 previous
    • Callmeishmael Callmeishmael on Oct 19, 2015

      @carlisimo Repeatability is sovereign when you're regulating, to be sure. I'd think that road testing would be mandatory just to backstop results and catch any outliers. I;m very surprised that with VW diesels standing head and shoulders above the others none of the regulators ran a road test just out of idle curiosity.

  • Alluster Alluster on Oct 19, 2015

    There is a way to fix this mess in a way that pleases all parties involved. Surely VW, employer of several hundred thousand can't be allowed to go bankrupt. Buying and destroying the affected cars will cause a bigger environmental damage than the pollution emitted by them. Shipping them to other countries will only move the pollution elsewhere. Retrofitting urea tanks will not please the owners. A better fix would be to estimate the combined NOX emitted by all affected vehicles over the compliance limit and find ways to lower an equivalent amount elsewhere. There are 20+ year old semis and 50 year old locomotives still in service. The companies owning them would love to replace these aging, maintenance money pits. VW could replace these tier 0 trucks, farm equipment, locomotives and construction equipment with tier 4 complaint machines. The companies would receive efficient replacements for free, lower maintenance costs, burn less diesel and lower NOX emissions by 95%. VW will pay for all of it but still cheaper than replacing 500,000 cars for free or pissing off owners by retrofitting their cars. The net emissions will be lower and the EPA wouldn't have to fine them anymore (maybe a small fine for lying) EPA's decisions even with the best intent are not without consequences. They banned EMD from selling locomotives unless they are tier 4 complaint. GE, the only manufacturer selling a tier 4, cannot make enough of them in a year and this has caused a massive locomotive shortage. Railroad companies were forced to bring back retired tier 0 locomotives into service that emit hundreds of times more pollutants than EMD's "non-compliant" tier 2 locomotives.

    • See 1 previous
    • RHD RHD on Oct 20, 2015

      Finally, a creative and intelligent post. Your idea is pragmatic and logical. VW will have to be punished, their car owners will have to be reimbursed, and the excessive and illegal pollution will have to be compensated for. It's a matter of equity, in the legal sense, while taking into consideration the need to not have innocent employees lose their livelihoods. So much more well thought out than the ubiquitious knee-jerk "lefty" bashing that attempts to pass as discourse.

  • Rudolph Rudolph on Oct 20, 2015

    Glad I did not upgrade from the 2002 1.9l TDI Golf I drive • Wish I still had the 1968½ M-B 220D with its 50+mpg (due to extremely heavy crankcase oil consumption after running engine blew a radiator hose in middle of a winter night) — I had to add oil more frequently than diesel fuel • Not too many years ago , some chap added an x-53 or x-71 GMC 2-cycle diesel to his ½T pickup truck • I’ve no idea of its particle output with the almost #2 heating oil of the era •

Next