By on March 4, 2010

TTAC’s attempt to correlate NHTSA data with sales to generate a UA complaint rate was the first, even if it was flawed. Others have now waded into the 776,482 lines of data available from NHTSA (anybody need something to do?), and while their efforts are admirable, they’re not yet good enough. UA rates by maker or brand fail to show how different the UA rate is for certain models, which is essential in tracking down the issue. The UA problem is not a make-specific problem generally, but a model specific one. More on that when we revisit that with updated stats.

NPR has come up with a data base that correlates UA by model year (MY), manufacturer, and the rate per 100k cars sold, from 1990 through 2009. It’s all here. I’ve taken a couple of key lines of that for the following chart, comparing Toyota’s rate to Ford’s, which has had a fairly consistently highest rate among the biggest manufacturers, and also picked the highest rate per maker for each year, not including any below five complaints due to the low margin of error:

It clearly shows that Toyota’s UA rate took a jump in the 2002 MY. Until we have some model specific info, we’ll have to guess as to why. It also shows that Ford’s average rate is almost as high as Toyota, but has dropped substantially in the last couple of years (we have a theory for that, coming in a later post). It’s also quite obvious that Toyota’s rate is not very high in comparison to the worst offenders.

[Update: Commentator david42 has made two additional charts from this data, showing accelerator events and all type of vehicle speed incidents. Thanks!]

Get the latest TTAC e-Newsletter!

Recommended

46 Comments on “NHTSA Data Dive 2: UA Rates 1990 – 2009 By Manufacturer [Updated With New Charts]...”


  • avatar
    Telegraph Road

    Here’s the answer about 2002 from the LA Times: “Toyota introduced electronic throttle controls in 2002 on certain Camry and Lexus models, and since that time consumer complaints to NHTSA about sudden acceleration have quadrupled for these models.”

  • avatar
    CarPerson

    A tip of the hat to TTAC for this:

    UA rates by maker or brand fail to show how different the UA rate is for certain models, which is essential in tracking down the issue.

    To my eyes, the NPR table throws everything into a bucket then trys to make a point with numbers that are essentially meaningless. Within each brand and sometimes even within the same model, the throttle design can be completely different.

    At the Senate Hearing, a chart was displayed showing the Toyota Camry trundled along at about 35 complaints of UA per year until 2003, when it jumped to 210 (a 6x increase) and slightly less the next year. The sales numbers and demographics were about the same through those years.

    It’s hard for me to believe that model year 2003 Toyota Camry drivers suddenly got stupid.

    In the Hearings, it was stated the 2003 MY was the year the Camry went to an all-electronic throttle.

    • 0 avatar
      midelectric

      I’d want to know if in addition to the change to electronic throttle control, were there changes to the dimensions and positioning of the pedal compared to the year before.

      That would help to isolate whether the addition of electronic throttle appears to cause such a spike in complaints.

    • 0 avatar
      Paul Niedermeyer

      The MY 2002 Camry was largely an all-new car. I’m pursuing two lines of inquiry: when did the Camry go to e-pedal, and did this new ’02 Camry have a pedal design that was much easier to entrap than the prior model.

    • 0 avatar
      Telegraph Road

      I think you may be referring to this slide:
      http://commerce.senate.gov/public/?a=Files.Serve&File_id=ddacce41-aed1-4e47-8c5c-4384bd7522ac
      which shows MY 2002 Camry beginning the spike.

    • 0 avatar
      CarPerson

      The chart that was shown at the Senate Hearing is at the link.

      The Senate chart shows about 35 UA claims but the second chart above shows over 50 in the years proceeding the electronic throttle, so there is a slight variation in how the data is being tallied.

      In verbal testimony it was stated 2003 was the electronic throttle changeover year but the Senate Hearing chart shows the changover year as 2002. Nonetheless, both show UA claims in the Toyota Camry shot up in the 2002-2003 time frame.

      What was not shown by the Senate chart but is evident here is that of three similar vehicles being operated by an arguably similar demographic, one of them, the Toyota Camry, is wildly over-represented and has been since 2002.

    • 0 avatar
      CarPerson

      @ Paul Niedermeyer
      I’m looking forward to your report on the 2001 vs. 2002 pedal design. I have a few comments…

      I’d like to see illustrations of the floorboard and firewall in profile with the pedal arc, to scale, above.

      I am expecting to see the 2001 has a greater clearance, but more importantly, the transistion from substantially parallel motion to substantially perpendicular motion between the pedal and firewall occurs before the clearance drops below 1/2 inch and at least 1/2 inch before the pedal reaches the end of travel.

      In the 2002, I’m expecting to see equal or less clearance, but more importantly, the pedal motion still has a large parallel component to it as the clearance drops to less than 1/2 inch or reaches the end of travel (hard contact with the carpet).

      The significance of this is that the “Pedal Entrappment” is related to parallel travel, the ability to wedge. Perpendicular travel, straight-on, lacks the 90-degree force to wedge. The arc of the pedal accross the floorboard and firewall will tell you how suseptable your design is to pedal entrappment.

      Push the accelertor pedal to the floor in a MY2002. If it makes a noticable sliding contact with the floor instead of dead straight on, the pedal arc length or point of rotation is screwed up in relation to the firewall angle. They have to intersect at a true 90 degrees to be completely free of a tendency to wedge.

      The transistion between the floorboard and firewall can be too far forward but the pedal arc should have been positioned to correct for this.

  • avatar
    jmo

    Here is the question I’d have – what was the thought process behind the engineering decision not to go with a brake application over ride for the throttle?

    Did some young Toyota engineer think of it and it was shot down, or did no one think of it?

    One other note – I wonder if the UA claims can also be attributed to rising bhp? SUA in a 4-speed AT 115bhp 1990 camry is something totally different than UA in a 2008 263bhp camry with a 6-spee AT.

    • 0 avatar
      psarhjinian

      You’d have to ask that question of several marques, because Toyota by far isn’t the only one not to do it.

      Part of it might be a desire to keep the code simple; note that the companies that do have it are mostly European or Euro-influenced, and European design philosophy seems to tend towards the complex.

    • 0 avatar
      ash78

      I posited on another site (semi-jokingly) that Toyota never built in a brake override because of the number of elderly customers they have…who often drive an automatic with two feet. As of a couple years ago, Toyota had the highest average buyer age among all mass market companies.

    • 0 avatar
      Robert.Walter

      Toyota clamed some weeks before the congressional hearings that the reason they had not introduced this feature was because they were still studying how to deal with starting a manual trans car on a hill … “still studying”, sounds much the “unverified prototype EDR reader” … but after things started blowing up, just as they said they will have 100 EDR readers availabe in the US by April (seemingly later backtracked – if I understood the testimony – to mid-2011), they seemed to produce a re-flash s/w in time to incorporate into the pedal field campaign.

  • avatar
    psarhjinian

    When did the NHTSA open it’s Complaints reporting system to the web?

    • 0 avatar
      1996MEdition

      Good question. As the reporting system is totally voluntary, it is not a random sampling and cannot be analysed with simple statistics. Also, stuck in my recent memory are NHTSA ads giving information on where to go on the web to report issues with your vehicle. When did those ads start? Would make an interesting study to see if complaints increased/decreased/or remained steady.

  • avatar
    Mr Carpenter

    Does anyone see “TOYOTA” on the right in any of the years as the “worst offender” ? Answer: NO. Yet there are some interesting things such as the seemingly sudden 250% increase between 2001 and 2002 model year cars; apparently this was when electronic throttle was added to the Camry, which makes up much of Toyota’s sales. Or, it could be a coincidence and simply down to the design change of the pedal to accommodate the electronic control, of course.

    But if you continue to look, the numbers get better again (except for 2007) under Toyota, and looking at the figures for Ford, they vary from year to year, as well. Of course, until 2009 (the numbers of which will be skewed by the massive hyperactive media blitz against Toyota), Ford and Toyota were not hugely different from year to year, except for 2002-2003 and 2007.

    Look at how much “worse” the stats are for the “highest make” compared to Toyota, though. Perhaps, for example, ALL 1999, 2000 and 2001 Volvos should be recalled, eh? Not to mention ALL 2003, 2004 and 2005 Jaguars. For that matter, look at 2002. Subaru?! Where’d that come from? Never re-appears on the “worst” list, though, do they?

    I still maintain that this is a witch-hunt against Toyota by the UAW and Government owned GM “fanboys” (and controllers) in Washington DC.

    • 0 avatar
      menjo

      All 1999, 2000, 2001 Volvos did get recalled for problems with the ETM (in most cases the cars stalled, but UA also happened).
      http://home.comcast.net/~donwillson/Introduction.html

  • avatar
    z4eva

    You must also look at higher-level complaint trends to see any true UA trends. For instance, did the total number of complaints per 100K cars change over this interval? Was the average driver getting older and, therefore, potentially more mistake-prone (think of the older folks who mistook their Audi accelerators for brake pedals)? Did the categorization of a UA complaint change — was that option added to the complaint form at any time, or did a “UA event” get redefined?

    This analysis is much, much too high level, and not just because it omits models.

  • avatar
    david42

    I took the NHTSA data and analyzed it by make, model, and year for incidents classified as “VEHICLE SPEED CONTROL:ACCELERATOR PEDAL”

    You can download my tables and charts here:
    http://www.sendspace.com/file/rs5wi5

    It’s interesting that there were two “good” years for the Camry (2006 and 2008), while 2007 was a disaster.

    I’ll soon post a comparison of Camry vs. Taurus vs. Accord.

  • avatar
    YotaCarFan

    I have a theory that may explain the 2001-2002 jump in SUA for Toyota.

    I had a 2001 and a 2002 Camry. Yes, the 2002 had the e-throttle. There was only one thing that seemed different to me driving-wise w.r.t. the 2002: The cruise control. With the 2001, when the cruise was engaged, the gas pedal was physically held down at the position it was at the moment cruise engagement occurred, by a servo pulling the throttle cable. If the driver took his foot off the pedal, the pedal remained stationary (partially depressed). The driver therefore had physical feedback that the gas pedal was “pressed”.

    With the 2002, the gas pedal was not held down physically by the cruise control system. If the driver took his foot off the pedal during cruise, the pedal would return to the “idle” (up) position. It took me a bit to get used to, b/c sometimes I’d take my foot away and then put it back and feel the pedal “up” and think the engine was idling; but, it was not, as cruise was engaged. Perhaps this difference in pedal operation during cruise control is leading to drivers to think the engine is at idle when it isn’t, and to misinterpret proper cruise control speed maintaining/acceleration as SUA?

    • 0 avatar
      Ion

      You should enter this to Edmunds I really think you have something.

      I use CC on a daily basis in my car and always used the pedal’s movement to gauge whether or not the car is accelerating too fast or too slow, CC would feel so disconnected to me without the pedal’s movement. The only problem though is that CC shuts off when you step on the brakes.

    • 0 avatar
      Robert.Walter

      @fan. A good and plausible theory, but it has to be qualified with the knowledge that some cable-actuated systems did a similar thing and some didn’t (this part I know with more confidence than what follows below). It is possiible I don’t accurately remember this long-distant memory, but I used to be aware of this pedal-CC relationship and notice it from time-to-time in vehicles.

      But then, like 20 years ago, for no particular reason (perhaps boredom or distraction), I seemed to have stopped checking for it or become desensitized to it.

      The memories are so dim, that I am no longer sure, but I think my ’89 Golf (w/Bosch KE-Jetronic) actually moved the pedal in proportion to the cruse-setting, so if you had your foot on the pedal, and moved the sliding switch to “+ speed” (or was it depress the button at the end of the stalk?? Ugh, memory!), the pedal would drop away from your foot.

      On the other hand, one of my parent’s cars (Ford?) let the pedal return to the full-out position…

      I was trying to recall the layout of these systems, and how the cables attached to the pedals, and I think I’ve seen pedals with double cables, and some with single cables, but I can’t remember the layout, or (at least as I type this 5 minutes before bed) imagine the kinematics that would require or make this possible.

      I had totally forgotton about this casual fascination until reading your comments, and, unfortunately, this is one of those rare times where I don’t have a solid memory of past events (perhaps my carbon-based internal EDR over-wrote this data due to limited memory or buffer-overflow…) … only thing I know, is that I think not all cable-attached pedals are created equal.

      I’m interested if others ever noticed such a phenomenon in their cars … (Btw, my 2003 Smart ForTwo has an ePedal, no brake overide, and when the cruise is on, and the foot is off, the epedals internal spring(s) return it to the 0° actuation position.)

  • avatar
    tced2

    What was going on with Volvo in 1999-2001? The rates are the highest on the chart for all makes and all years – about double the average rate. Why didn’t we hear about Volvo from the experts in the media? (And Volvo was part of Ford during this time).

    • 0 avatar
      Gregg

      Ford bought Volvo in ’99. Chances of them having any design/engineering input into even the 2002 Volvo is slim to none.

    • 0 avatar
      menjo

      They used an ETM that almost universally failed after a certain time.
      http://home.comcast.net/~donwillson/

    • 0 avatar
      tced2

      After reading the webpages about the failure of the ETM, I would suspect that the failure of the Volvo ETM in the years 1999-2001 commonly caused the engine to not be run like the accelerator was fully pressed. Hence you would not get out-of-control acceleration complaints. And not much media coverage. And no Congressional hearings.

  • avatar
    srogers

    Did foot wear fashion for codgers change in 2002?

  • avatar
    david42

    Here’s a chart that compares this type of incident (“VEHICLE SPEED CONTROL:ACCELERATOR PEDAL”) for Camry, Accord, and Taurus:
    http://www.sendspace.com/file/220kgt

    Camry is clearly worse than the other two, and all three were sold in roughly equal numbers. It’s interesting that these problems significantly DECLINED for Taurus and Accord over time. Perhaps an accelerator-override was implemented in those models at that time???

    • 0 avatar
      Runfromcheney

      IDK about the Accord, but it is hard to explain the Taurus’ decline, as the graph ends at 06, and every Taurus built up until 2007 had a cable driven accelerator. It didn’t go to an e-throttle until 2008.

  • avatar
    ash78

    If we go back to the Audi UA issue from the 80s, I’m wondering if the issue carries forward to the newer years–also dominated by Euro marques among the “worst offenders.”

    My hypothesis is pedal placement, just like the Audi UA debacle.

  • avatar
    david42

    Last post: here’s a chart that shows the number of incidents (“VEHICLE SPEED CONTROL:ACCELERATOR PEDAL”) across ALL models.
    http://www.sendspace.com/file/swwmdq

    There is a huge spike for the middle of the year range. Seems pretty bizarre: why would vehicles built in those particular years be so much worse than those that came before AND after?

    Could it be that older vehicles are no longer on the road, and therefore didn’t make it into the reporting system, combined with the fact that newer vehicles may have accelerator-overrides? I don’t know, but it’s an interesting question….

  • avatar
    csliwins

    What roll might the increased prevalence of driving while using a cell phone have on this?

    • 0 avatar
      Runfromcheney

      Good point. The graph starts skyrocketing in the early to mid 1990s, where phones in cars became commonplace. I remember reading a Jean Lindamood editorial in a 1991 issue of Automobile where she mentioned using a Pontiac Bonneville with a car phone for a week, and her driving suffered noticeably when she used the car phone. So that is a reasonable explanation.

  • avatar
    david42

    OK… one more! There are other kinds of “VEHICLE SPEED CONTROL” incidents, not just accelerator pedals. I was curious to see if including the others have any impact. There are far more cases (for all types of cars), but the pattern is the same.

    http://www.sendspace.com/file/2cjgk7

  • avatar
    jmo

    they had not introduced this feature was because they were still studying how to deal with starting a manual trans car on a hill

    Question: Are ECUs on manual transmission cars the same as autos?

    If anyone has an DSG GTI try and use the gas and brake and see what happens. On manual GTIs the application of the brake does not interfear with the throttle. But, obviously SUA is not a problem in a car with a clutch.

    • 0 avatar
      tced2

      I have an Acura TSX with drive-by-wire and manual transmission. Pressing the brake pedal does not appear to have any effect on the accelerator controlling engine RPMs.

    • 0 avatar
      criminalenterprise

      Check to see when the car is in gear and under way if you can apply brake and throttle simultaneously.

    • 0 avatar
      Zarf

      I thought I read that the GTI with DSG actually has a launch control mechanism where it requires you to press the break and gas simultaneously to maximize acceleration…if you wanted to do that sort of thing.

    • 0 avatar
      tced2

      Acura TSX with drive-by-wire and manual transmission.
      Driving in 3rd or 4th gear and pressing the brake doesn’t seem to alter the engine RPMs. I don’t think the engine computer “knows” anything about the what gear you’re in.

  • avatar
    JohnAZ

    Although it is useful to compare Ford(4.77) and Toyota(6.43) back to 1990, I think it also useful to compare them over two other periods.

    Comparing them since Toyota went to their ETCS system shows Ford at
    6.06 and Toyota at 12.08 (almost double).

    Also comparing them since Alan Mulally took charge shows Ford at 2.77, which is a 54% improvement, and Toyota at 9.93 which is more than 358% more than Ford and only a 19% improvement.

    Aren’t stats great?

    • 0 avatar
      ihatetrees

      Aren’t stats great?

      In this case, stats would be great only if the data were downloaded from every vehicles’ ECU and transmitted to the NHTSA.

      But, in this case, I’d say stats are mediocre. This data is self-reported. Via the web, phone, and mail, with a heavy bias toward the web.

      I’m certain there are notable statistical differences in the web usage rates among purchasers of different makes. Along with the cell phone theories noted above, there’s a lot of smoke to filter through here.

  • avatar
    Z71_Silvy

    So…Ford is either 1st or 2nd 19 out of the past 20 years and that gets over looked by the media…over the past 10 years…Ford ON AVERAGE has has 211 complaints per year while Toyota has had–ON AVERAGE–229.

    And yet the media is only focusing on Toyota and not Ford. Why? Am I to assume that a difference of 18 vehicles on average makes Ford completely infallible? And meanwhile, Ford gets a free pass when the braking system fails on the mediocre Fusion/Milan Hybrid…and only issues a fix to the customers when Consumer Reports calls them on it.

    Our media is the worst in the world.

    • 0 avatar
      ihatetrees

      Our media is the worst in the world.

      +1.
      What can you do? J-skool grads tend toward the unscientific and statistically retarded demographic. But the good news is that the current recession will kill off a lot of this media. Better yet, J-skool grads make excellent wait-staff at higher end restaurants. It may reverse the decline in dining service I’ve noted in the last decade.

    • 0 avatar
      criminalenterprise

      I wonder if Ford events are of similar duration or lethality as those experienced by Toyota and Lexus.

  • avatar
    red60r

    A big difference in pedal behavior happens with Drive By Wire systems on cruise control: the old way used an electric or vacuum-operated servo to control the throttle, which was mechanically linked to the accelerator pedal. As the car traveled up up and down hills, the accelerator would magically move as the system added or reduced power demands. With DBW, there is no physical connection between pedal and engine, nor is any needed. The servo on the throttle operates under computer control, so the driver doesn’t feel any invisible foot operating the pedal for him. If the system works as intended (this may be where Toyota’s problem arises…) there is less chance of the throttle hanging up in a particular position due to a sticky cable or linkage.

  • avatar
    Mr Carpenter

    I never thought of cell phone users being distracted, but it could be part of the answer. I have lost count of how many times knuckle dragging sub moronic imbeciles have nearly driven into me while incessantly blabbing on their damndable cell phones.

    I suspect it’d be near impossible to figure out rates of UA parsed out by levels of cell phone use while driving.

    One thing I have noticed; people younger than about 25 get straight in my face and defend their “right” to blab incessantly on their phones when I speak about how bad a practice it is. They’re not even willing to pretend that they don’t do it and lie. Quite the opposite; they try to put me on the defensive for stopping THEM from what THEY want to do, even it if can demonstrably be proven that it’s a dangerous activity.

    Blatant selfishness. If I didn’t know better, I’d say they were human beings acting in the way that most do…. (sarcasm).


Back to TopLeave a Reply

You must be logged in to post a comment.

Recent Comments

  • speedlaw: Parts replacement. One source of repairs. Whompy Wheels. Pass
  • thelaine: It is a great credit to Musk’s salesmanship that Tesla still exists. Without the cult, different...
  • labelnerd: Even if I was interested in a Tesla and had the ability to buy one, I would not due to their ineptitude...
  • formula m: It feels like a 2019 Ram Sport, Canadian availability only.
  • Corey Lewis: A longer, relaxed trip would’ve been nice. But the time off work was short notice, so I respected...

New Car Research

Get a Free Dealer Quote

Staff

  • Contributors

  • Timothy Cain, Canada
  • Matthew Guy, Canada
  • Ronnie Schreiber, United States
  • Bozi Tatarevic, United States
  • Chris Tonn, United States
  • Corey Lewis, United States
  • Mark Baruth, United States
  • Moderators

  • Adam Tonge, United States
  • Corey Lewis, United States