The government is changing. After some 32 years of crash testing cars and trucks in much the same manner, the National Highway Traffic Safety Administration (NHTSA) is finally updating its crash test procedures. You know all those "five-star ratings" you see touted by one carmaker or another? They're going to be a little harder to come by with the new procedures. Here's why.
More Dummies Involved
One of the most significant changes to the NHTSA's procedures is the addition of a smaller, female dummy in the car. Hard to believe, but the NHTSA has used nothing but an average-size male dummy to approximate the probability of injuries in a crash. Last time we checked, there are a few women who have taken up driving. It was inevitable after they got the right to vote and everything.
According to the NHTSA Administrator David Strickland, "We recognize that this didn't properly capture a significant portion of the American populace. We added the small female dummy to capture how a crash would impact them in terms of injuries."
Additional updates were made to the standard male dummy. More injury sensors were installed to give the researchers more accurate feedback on the severity and type of impacts suffered during the various collisions.
A New Crash
The three crash tests used before — front, side and rollover — continue unchanged with the new procedures. The new twist in the equation comes in the form of a side pole impact test. It's designed to simulate hitting a telephone pole or other such narrow, stationary object.
To approximate such a disaster, the NHTSA pulls the car sideways at 20 mph until it impacts a 25cm pole at the driver's door. Curiously, the NHTSA uses the female dummy for this test. There was no explanation why, but we figure the agency wants to get maximum use out of its fancy new dummies. The NHTSA also did not choose to address why it sticks with its traditional front impact tests while outside competitors like the Insurance Institute for Highway Safety (IIHS) use an offset crash that better simulates real-world accidents.
Speaking of fancy, the NHTSA is also going to start tracking the availability of electronic safety devices as well. Three features — electronic stability control, lane-departure warning systems and forward-collision warning systems — will now be shown on the window sticker. The presence of such systems won't affect a vehicle's overall crash score.
What To Expect
So what do these changes mean to you? Well, for one, you won't be seeing quite as many five-star ratings as before. In the NHTSA's first round of testing under the new guidelines, only two out of the 34 vehicles tested earned a full five stars. Keep in mind, however, that 2011 and later test scores won't be compatible with the earlier scoring system, so there's no way of comparing the two systems. The NHTSA will also be introducing an overall rating that combines the results of all the individual crash tests into one rating.
You should also look for a revised window sticker that shows the new ratings. The design hasn't been finalized yet because it also involves the EPA's new emissions rating, but we expect to see the changes made by next year. The goal with the new look is supposed to be clearer, more usable information for consumers. We'll have to wait and see if a government agency can actually come through on that promise.
Add A Comment »
blurglide says:
04:13 PM, 11/01/2010
I'm not sure why the comments I added earlier were deleted, so here's a repost...
You know what would be really helpful: A rating with a little more fidelity than 1-5 stars. The best 5 star cars have less than half the injury rate of the worst 5 star cars. The rating should also be weighted by the likelihood & typical severity of certain types of accidents. For example, only a couple of percent of people who get rear-ended die, while front, side, and rollover impacts are far more dangerous. Finally, a barrier crash test is only useful if you hit a barrier or a car the same size as yours. More often than not, you're hitting a vehicle of differing mass. Data should be mined to find out how often vehicles are in single car vs. multi-car accidents, and a weighting factor should be applied to the rating to take mass into account. All that together should be given a single number representing your likelihood of being killed or seriously injured in a crash, starting at a theoretical best of "0" for something impossible to die in, and going up from there.
Informedforlife.org tries to do something very similar to this, but it's just some guy crunching numbers. He obviously doesn't have the resources to crash cars, so he has to interpret the data from NHTSA and IIHS crash test. The data is often incomplete, or of insufficient resolution, resulting in less than consistent results. Still it's the best system available at the moment.
Finally, instead of just measuring rollover risk, perhaps an overall accident avoidance test could be put together, putting emphasis on braking, handling during swerving, and ability to "thread the needle", for example. This combined with the electronic safety features included, could be rolled into an accident avoidance score. Of course tires & weather conditions would play a big role in this test, so it'd be difficult to get consistent meaningful results.
Oh yeah- one more thing. Instead of crashing the car into a flat barrier, perhaps car registration data could be mined to find the median bumper height for vehicles on the road. The barrier could then have a bumper-shaped protrusion on it, sticking out 6" or a foot.
This would encourage car makers to design their cars to put their bumpers at this height to improve their crash test score without actually levying a REQUIREMENT for them to do so. This would improve safety, but still allow carmakers to keep producing big trucks and low-slung sports cars, and we wouldn't wind up with big-nosed monstrosities like are coming out of Europe lately.
blurglide says:
06:46 AM, 11/01/2010
Oh yeah- one more thing. Instead of crashing the car into a flat barrier, perhaps car registration data could be mined to find the median bumper height for vehicles on the road. The barrier could then have a bumper-shaped protrusion on it, sticking out 6" or a foot.
This would encourage car makers to design their cars to put their bumpers at this height to improve their crash test score without actually levying a REQUIREMENT for them to do so. This would improve safety, but still allow carmakers to keep producing big trucks and low-slung sports cars, and we wouldn't wind up with big-nosed monstrosities like are coming out of Europe lately.
blurglide says:
06:38 AM, 11/01/2010
You know what would be really helpful: A rating with a little more fidelity than 1-5 stars. The best 5 star cars have less than half the injury rate of the worst 5 star cars. The rating should also be weighted by the likelihood & typical severity of certain types of accidents. For example, only a couple of percent of people who get rear-ended die, while front, side, and rollover impacts are far more dangerous. Finally, a barrier crash test is only useful if you hit a barrier or a car the same size as yours. More often than not, you're hitting a vehicle of differing mass. Data should be mined to find out how often vehicles are in single car vs. multi-car accidents, and a weighting factor should be applied to the rating to take mass into account. All that together should be given a single number representing your likelihood of being killed or seriously injured in a crash, starting at a theoretical best of "0" for something impossible to die in, and going up from there.
Informedforlife.org tries to do something very similar to this, but it's just some guy crunching numbers. He obviously doesn't have the resources to crash cars, so he has to interpret the data from NHTSA and IIHS crash test. The data is often incomplete, or of insufficient resolution, resulting in less than consistent results. Still it's the best system available at the moment.
Finally, instead of just measuring rollover risk, perhaps an overall accident avoidance test could be put together, putting emphasis on braking, handling during swerving, and ability to "thread the needle", for example. This combined with the electronic safety features included, could be rolled into an accident avoidance score. Of course tires & weather conditions would play a big role in this test, so it'd be difficult to get consistent meaningful results.
farvy says:
09:05 AM, 10/22/2010
NHTSA didn't need to come up with the side pole impact test. They can just give Edmunds IL staff the cars to use in their parking garage.
calspecial68 says:
10:22 PM, 10/21/2010
@ shouldermonkey: I mean, it only ran into an immovable wall. Even a heavy duty truck is gonna look weak against that thing.
I like the idea of stricter safety requirements and new standards, but I also agree with rpvitiello. The US and EU should try and adopt a compromise of both systems and save everyone the headache and the money, but of course no one will ever do this. Also, what about better driver's education programs in this country?? These programs are so old-fashioned. To skip the hassle of designing a new system of education, Bob Bondurant should just implement his stuff into it all. How about it guys? Everyone will be a better driver that way.
rpvitiello says:
09:20 PM, 10/21/2010
I just don't understand why the US and the EU cant just use the same dam test! why waist money crashing a car all over again to perform a very similar test that was already done! if they used the same system, one group (ncap and nhtsa) could test some cars and the other could do the rest. rather than have 2 tests on the same car and none on others (since not all cars get tested) they would have more cars tested for the same money.
compressor says:
10:30 AM, 10/21/2010
Two Questions.
1. Will the female dummy sit 6in from the steering wheel with an anxious look on her face? (joking btw, nevertheless my sexist flame suit is on )
2. Can we now expect even less outward visibility and higher vehicle weights? not joking.
frank908 says:
07:01 AM, 10/21/2010
@ant14
I've traveled a lot of Europe and there's a good amount of fat asses there too. It's not just America anymore.
dg0472 says:
06:41 AM, 10/21/2010
It is a Solstice; under the old system it got four stars in front and side. From the picture, it appears to be a reasonably good performance: the dummy's head is dead-center in the airbag and all of the collapse has occurred fore of the firewall. Unfortunately the IIHS never tested it and Euro NCAP never tested the Opel GT. I couldn't find any crash tests for the Daewoo G2X either. But there's nothing in that picture to indicate the Solstice/Sky are an unreasonable crash performer for a vehicle their size.
shouldermonkey says:
05:44 AM, 10/21/2010
Is that a Pontiac Solstice? To think I wanted one.