Drive-by-wire backlash?
Concerns about software bugs may be creating a backlash against x-by-wire systems, based on commentary appearing in mainstream media. For example, in today’s Washington Post, Eugene Robinson waxes nostalgic for what cars once were: “Automobiles used to be mechanical devices,” composed of an air cleaner, a distributor, and “other parts that backyard mechanics of a certain age will fondly remember.” He continues, “Now, they are collections of mechanical parts that are told what to do by computers…. When something goes wrong with a car, you don’t start by opening the hood and unbolting pieces from the engine one at a time, the way we used to. You plug in a reader device and ask the vehicle what its problem is.”
Holman Jenkins in the Wall Street Journal offers a similar take. He says that lacking hard data, we can attribute some, but not all, of Toyota’s sudden-acceleration problems to driver error, floor mats, or sticky pedals, and he asks, “Of those incidents that remain, must we assume an electronic glitch as the default explanation even if it can’t be proved?” His answer: “Perhaps inevitably, which is a problem for the whole industry.”
Jenkins claims that the price of a new car (with electronics and software representing up to 40% of its value) “will surge as the cost of testing and retesting vehicle electronics now skyrockets” and as carmakers build into prices legal liability protection for executives and shareholders.
The problems will only get worse, Jenkins says, with “the forthcoming Chevy Volt and other Obamamobiles,” in which “electronic complexity will take another quantum leap. We’re not just talking about more sensors, algorithms and look-up tables for the purpose of optimizing emissions, but to coordinate two completely different power systems, electric and gas-powered.”
Both Robinson and Jenkins note that this is an industry problem, not a problem for a single manufacturer. Says Robinson, “Toyota’s competitors should go easy on the gloating. Their cars are fly-by-wire, too, and thus equally at the mercy of information-age technology.” Adds Jenkins, “GM, Ford, Volkswagen, and other competitors may be indulging a certain satisfaction right now at Toyota’s troubles. Perhaps they shouldn’t.”
There have long been parodies about what cars would be like if personal-computer companies built them. Here’s one from the Windows 95 era. Now, many people might begin to assume that the parodies are becoming too close to the truth. Concludes Robinson in the Post, “Raise your hand if you think it’s a great idea to make our cars precisely as dependable and problem-free as, gulp, our personal computers.”
Follow me on Twitter: www.twitter.com/Rick_editor.
Roger K commented:
When designed properly, drive-by-wire saves gas due to lighter weight and can be as good or better than mechanicals. On top of it, we are denying those oil rich countries who hated us from getting our money and use them to kill us!
WT commented:
If you read the USA Today article cited in the comment from Martin Rowe below, there is an excellent comment posted there by an automotive engineer ('HillFarm') in which he argues there is no way EMI could be causing the SUA events. I respect his opinion and the arguments appear sound, however, there are too many unexplainable incidents of SUA in late-model Toyotas to completely dismiss this as just floor mats and a few sticky pedals. Every Toyota owner and automotive engineer should read the recent report by SRS:
'www.safetyresearch.net/2010/02/05/toyota-sudden-acceleration-the-full-report-from-safety-research-strategies/'
mad dog commented:
Drive-by-wire is subject to the same problems aircraft have with fly-by-wire. Engineers should know better than to place one set of command wiring next to another (no damage tolerance), and not have a manual, mechanical back-up. Aviation "insiders" often assume that because the military has had some success with fly-by-wire that the "system" has been perfected. Each system is individual and what may work for one model, may not work in another. Excellent design, superior hardware and software make a vehicle successful, however robustness (weight/damage tolerance) bring the vehicle home to run another day.
Mike commented:
I do find it so typical that toyota has a problem and the comment is GM/Ford should not gloat, it could be them next. If the problem were in a GM/Ford car, the comment would be "Don't buy american cars, they are crap and got what they deserved." Maybe op-ed pieces like this might want to ask it this way. Is Japanese quality real or an illusion? Is it worth the premium paid? Resale data is trending towards no its not real.
Miskit commented:
The computer metaphor is only relevent to PCs. My Mac never crashes. The secret is efficient and well designed software that isn't a hodge-podge of disaster code compliled into a crash waiting to happen.
fuji commented:
As some have stated, attributing some of the problems to some of the assumed causes will account for some of the incidents; the other instances of apparent failures cannot, and possibly will never be traced: what about the guy in the vehicle next to you, who stats talking to his friend on an illegal 1000 Watt linear amplifier for his CB radio? i'm sure all of that RF would have to do something to any electronic device within proximity... for myself, i like my older cars, that still have cables for throttles, and mechanical connections for my steering. one thing i have learned over the years - newer does not mean better!
DrX commented:
Adding safety contingencies to the software seems like a no brainer and something that should have been addressed if a true DFMEA was performed on this vehicle. Having the brake override the accelerator information seems like something which should have been included from day 1 and mighth have avoided this whole situation.
Steve P commented:
"Obamamobiles"? Please.
Face it, becuase of the demands from customers and government, vehicles are more complex than they used to be.
I'm sure those crazy horseless carriages had the same issues back in the day. Why, all you used to do was feed them oats...
Not a GOP commented:
Do all the ultra conservatives sit at their terminals all day looking for any issue to attack Obama? Do they have any idea of how *STUPID* they look? What a waste of time. Weren't they the ones that continuously harped on 'a true patriot supports the president'? Clearly these guys are all from DumbF*ckIstan.
John commented:
Obama's communist comrades have spotlighted the union's enemy - foreign manufacturers, because with the democrats everything from the Haiti earthquake to the common flu is a political issue. Fortunately the Barrak Hussein Obamites don't have the intellectual horsepower to match their evil ambitions.
TomFromNY commented:
I think the terminology "Chevy Volt and other Obamamobiles" is inaccurate and mean spirited. The Volt's design, as that of the Prius and Tesla have been years in the making. PRE OBAAMA. Proper SW design requires extremely careful design specifications, design of a thorough test scenario and, above all, attention to detail, not just stringing together commands in order to make something work. It also requires continually asking one's self what might go wrong.
Lee M commented:
I "fondly" remember having to manually adjust the valve lifter clearance on VW bugs every 3 to 5 thousand miles to avoid burning the valves. No self adjusting hydraulic lifters for the purists.
Of course over the long term manually misadjusted or neglected lifters burned a lot more valves than the occasional stuck hydraulic lifter.
What really matters in the end is the overall system reliability no matter what technology is used for the implementation.
Just because a car requires a computer hookup to diagnose a problem does not mean it is harder to mantain.
For example good engine controllers capture details on intermitent faults that would otherwise be very hard or impossible to find.
Martin Rowe commented:
USA Today reported that it could be an EMI problem.
www.usatoday.com/money/autos/2010-02-03-electromagnetic03_ST_N.htm?csp=34
Ernie commented:
Even worse problems will arise with ROHS chips and boards in drive by wire systems. Tin whiskers make these systems unreliable. How would you like to have your drive by wire steering reboot during a left turn!!!
