By on March 18, 2010

Here’s (perhaps) the finale of David’s remarkable data diving: a full chart showing all makes and models sold from 1995 through 2008, with their rates of reported UA incidents to the NHTSA. To make the findings easier to interpret, David has adjusted all the results as a relationship to the same year average, rather than just the raw results. This really highlights those vehicles with higher than average rates of reported UA.

The table has been inserted full after the jump, as well as our commentary, but if you want to access the excel file in its entirety, it’s here.

David:

This chart lets us directly compare any model/year, because it controls for the number of each model/year vehicle sold, and the number of years it has been on the road.  (We want to control for the number of years on the road, because older cars would generate more complaints simply because they’ve been around longer, not necessarily because they’re more prone to UA.)

Here’s an example of how to read the chart, comparing the 1998 Audi A4 with the 1999 Audi A6.

  • The 1998 A4 had 3 times as many UA incidents as the average 1998 vehicle.
  • The 1999 A6 had 6 times as many UA incidents as the average 1999 vehicle.
  • Therefore, we can say that the 1999 A6 is more than twice as likely to have a UA incident as a 1998 A4 (6x is twice as big as 3x).
  • This comparison is valid, even though there were more A4s sold than A6s, and even though the A4s have been around for an extra year.  These factors are controlled when calculating the numbers shown in the chart.

Some observations about Toyotas and other vehicles:

  • Camrys are consistently worse than average for 2002-2005 model years.  And the 2007 model has among the worst values for popular vehicles.  But 2006 and 2008 Camrys are almost incident-free.  The Lexus ES has the exact same pattern.  So it appears that the recall (which covers 2007-2010 Camrys) does not target some of the most affected cars (2002-2005).  Many of the cars which ARE recalled don’t have a significant UA problem (2008 and 2009–and 2009 has so few problems that it doesn’t even show up on the chart!).
  • Toyota has a problem across many of its vehicles, not just Camrys. It’s the only manufacturer where so many model lines have problems after 2002.
  • General Motors and Honda are nearly immune to UA issues.  GM’s success in preventing UA is surely meaningful, given that it’s a full-line manufacturer with models known to be preferred by older drivers.
  • Fords were quite problematic until 2002.  Cars from the 2003 or later model years are much less problematic.
  • There are many other model/years with incident spikes (e.g,. Volvos and Jaguars), but these are generally much lower-volume vehicles.  Since we don’t have much context about their UA history, it’s hard to say if the spikes are meaningful.

Paul:

What seems to stand out is numerous incidents of higher than average UA rates for vehicles just after a new generation is introduced, or the opposite: a dramatic reduction of a high UA car after it is replaced. Some examples of that are:

Ford Expedition: drops off after new ’03 model. Chrysler minivans: spike in UA after new ’96 model introduced. Ford Focus: spike after new model introduced in 2000. Suzuki Grand Vitara/Verona high UA rates drop after new model introduced in ’06.

What is clear to both of us is that issues with pedal location, size and relationship to their surroundings may be a key factor in these changes. This warrants further inquiry.

Get the latest TTAC e-Newsletter!

15 Comments on “NHTSA Data Dive #5: UA Incident Rate By Models 1995-2008...”


  • avatar
    Sandy A

    Thank you for the data and analysis. It’s good work.

    When I mentioned cosmic radiation as one possible reason for sudden acceleration some people compared it to divine intervention. Here is a story that appeared recently on cosmic radiation and Toyota’s sudden acceleration problem.

    http://www.1011now.com/home/headlines/88312092.html

  • avatar
    jmo

    So, VW Passat 2006….

    Nothing in 2005 nothing 2007 – what was up with the 2006? I know a new model came out in 2006, did they notice a problem and put in a code fix for 2007?

    Also, 1998 Passat, A4 and A6 were all very similar and VW has no incidence and Audi has 3? I’m not understanding what the difference could be…

    • 0 avatar
      wsn

      Maybe they have exactly 5 incidents in 2005 and 5 in 2007. Somehow the author thought 5 is a magic number that shouldn’t be included and should be altered to 0.

  • avatar

    Interesting. Would be even more user-friendly if non-production years were just blank or dashes rather than zeros. I wonder if the high initial prius number is because they don’t really seem “on” until you hit the accelerator?

  • avatar
    wsn

    Why only show models with 5+ UA cases?

    Why put 0.0 there instead of “inconclusive” for those cases? 0.0 = 0 report, which is a lie.

    All VW models show 0.0 in 2009, which can’t be true, as by brand, they have the highest rate.

    Don’t tell me the “statistically insignificant” crap. VW sell 200k units per year, which is a statistically significant sample.

  • avatar
    LennyZ

    Great data. Of course it leads to more questions like how many models have the drive by wire pedals as opposed to mechanical linkage. What year in the model’s history did drive by wire take over mechanical linkage?

  • avatar
    psarhjinian

    I am really starting to think this is an ergonomic or kinesiologic issue and not a mechanical or electronic one. That it’s so scattershot within manufacturers and even among shared platforms is fascinating.

    I would love to see driver height, leg and foot dimensions in all this; I’d also like to see footwell dimensions for vehicles on opposite poles (eg, the Panthers and ACV30+ Toyotas versus, say, the GM W-Bodies or Ford F-Series) Heck, I’d love to get a dummy into these vehicles and see what happens with regards to leg and foot movement.

    Toyotas seem to be short-driver friendly vehicles, from my experience. With few exceptions (Sienna, post-07 Tundra, Lexus GS/LS, Echo) I’ve had real trouble getting comfortable in them: short seat cushions, funny wheel placement, etc. I’ve found certain Fords to be cramped as well. I’ve never found that in Chrysler or Nissan’s products (GM has roof height trouble, but lower body dimensions don’t ring a bell). I’m just guessing, here.

    It doesn’t take much to set a driver’s leg/ankle/foot at a suboptimal position. Assuming that vehicles subject to SUA are driven by older (and usually shorter) drivers, there’s the some potential correlation, here.

    The other possibility is something like pedal feel, feedback and/or articulation. Again, I really wish I had some of my wife’s kinesio gear (and the time) to look at this.

    Another point is that ergonomic issues might not be a problem of their own, but they might be an aggravating factor when coupled with, eg, friction problems, mat entrapment and/or over-agressive throttle tip-in.

    • 0 avatar
      Patrickj

      You may well be right. Driving (and foot) positions vary drastically between vehicles. I can’t get comfortable in any Honda, including the Pilot, while I’m marginal in Toyotas. I’m favoring domestics as I get older.

      Drastic mistakes in vehicle purchases due to ergonomics are common. It is a common cause of early trades.

      Perfectly reasonable that UA incidents may be mainly an ergonomics/human factors issue.

  • avatar
    Dynamic88

    I’m not seeing much reason to blame driver height (or lack thereof). How is Toyota more “short friendly” than Honda? Or Mitsubishi?

    And certainly Volvo is as Tall friendly as GMC.

    • 0 avatar
      psarhjinian

      I can’t really prove it without photos, but have a quick read of Consumer Reports’ road tests for just about every Toyota model: across the board you’ll see complains about short seat cushions and awkward steering wheel reaches. Honda models, by comparison, has made Consumer Reports’ list of good cars for very short and very tall drivers.

      Volvo did have an electronic issue in the model years noted, true.

  • avatar
    210delray

    I know you said that this chart compensates for sales volume, but I also know that the 2006 and 2008 model years were exceedingly short for the Camry and likely the ES350 as well. Each model “year” (2006 and 2008) lasted only about 6 months, because the redesigned 2007s came out in March 2006, and the warmed-over 2009s came out in January 2009.

  • avatar
    Nicholas Weaver

    I think a huge problem however is noise: These “UA” events are very rare, and when you slice it up so finely, noise alone will make a difference, just one granny who mistakes the gas for the brake can screw up an entire year’s worth of data for a particular model.

  • avatar
    mpresley

    Is it just me, or is there something creepy about the people in “Their Next Car Will Be A Toyota” ads?

  • avatar
    Robert Schwartz

    I wear 13 2E shoes. I have noticed that I have a lot more problems with pedal placement — usually hitting both at once — on my wife’s Toyota, than on my Accord.

    I have often wondered why movable pedals have not been more popular.

  • avatar
    CatFan78

    Probably 99% of SUA complaints are driver error.

    Today NHTSA said NY Prius was driver error. California Prius was proven to be a hoax earlier this week.

    Toyota sales up 40-50% this month. Oh what a feeling !


Back to TopLeave a Reply

You must be logged in to post a comment.

Subscribe without commenting

Recent Comments

New Car Research

Get a Free Dealer Quote

Staff

  • Contributing Writers

  • Jack Baruth, United States
  • Brendan McAleer, Canada
  • Marcelo De Vasconcellos, Brazil
  • Vojta Dobes, Czech Republic
  • Matthias Gasnier, Australia
  • W. Christian 'Mental' Ward, Abu Dhabi
  • Mark Stevenson, Canada
  • Cameron Aubernon, United States
  • J Emerson, United States