Monday, July 23, 2012

Cars with safety aids are involved in more accidents.

Regular readers of my site and/or blog will know I have particular place in Hell reserved for pointless in-car gadgets like auto-lights, auto-wipers, auto-park, lane-departure warnings, blind-spot warnings etc. The list of these things goes on and on and I've long argued that when you fill cars with this stuff, it lulls the drivers into a false sense of security. The end result shouldn't come as a surprise to anyone then : cars fitted with lane-departure warning systems are involved in more accidents than those without. Cold. Hard. Facts.
The Highway Loss Data Institute (a division of the Institute for Highway Safety) discovered that lane departure indicators are doing more harm than good. They don't go into a lot of detail about why lane departure warning signals correlate with more wrecks, but the numbers speak for themselves.
Lane departure indicator-equipped Buick cars recorded a four percent increase in collision claims and a slightly higher rate of increase for property damage claims. Mercedes brought up the rear with a five percent increase in collision claims and a more than ten percent increase in property damage claims. Other manufacturers were scattered between the two with only one manufacturer - Volvo - bucking the trend.
The study results stand in stark contrast to researchers' initial predictions about how well these new safety gadgets would work. In 2008 the IIHS admitted that most of these things only partially work but at the same time boldly claimed that lane departure systems would prevent 10,345 fatal accidents per year (article). 4 years on, not so much.
Consider for a moment the arguments being floated in Europe right now for AEB - automated emergency braking. The exact same claims are being made right now for making AEB mandatory on all cars from 2014.
The HLDI won't speculate as to why the facts are the way they are, but it would seem pretty obvious. When you put crap like this in cars, it leads drivers to start to rely on it, so instead of actually paying attention when driving, they rely on the car to tell them when they're leaving the travel lane. In the same way that cars with blindspot warning devices are involved in more rear-end collisions (2008 report mentioned above) because people now believe they don't have to use their mirrors or look over their shoulders any more.
If governments want safer roads, they need to mandate getting rid of all this crap in cars and concentrate on driver training, testing and re-testing. If we continue the march towards more and more driver "aids", and ultimately auto-drive cars, the accident rates will only continue to march upwards as we dumb down the responsibility for driving.
(pic from Jalopnik)

Also : the competition is still running to win a BlueDriver wireless OBD2 scanner.Win a BlueDriver


Anonymous said...

I have no doubt that a lot of these technologies are oversold, and I also don't like the idea of less driver engagement with the task of driving, but there is an alternative explanation you have to consider in these situations.

The old internet saw, "Correlation does not equal Causation" applies here in spades. Are vehicles with these technologies perhaps more likely to appeal to those with diminished reflexes/eyesight/memory, who could be expected to have a higher accident rate anyway?

Chris said...

Good point - I'd not considered that. That would make an interesting study - is there a correlation between the age of people buying these cars and the accident rates with and without these technologies?

117527760007277806046 said...

I can very well imagine a correlation between the driver age and the people buying the systems. For example the average age of a BMW 7 series customer is much higher than the average age of a 1 or 3 series BMW. Most driver assistance systems are sold (and bought) initially only in the more expensive cars, so a correlation is certainly possible.

On the other hand I think the biggest problem now is that the people don't understand the limits of the systems. People are used to the fact that computers can process data much faster than humans and expect them to perform better in most situations. However when we talk about image processing for example it is not that easy. I'm convinced that if you better understand what the system can and cannot it can be very useful. Because there are cases where automated systems are much better than a human - you should just be able to recognize them :) And I think people will learn with the time.

P.S. I'm really biased since I'm deeply involved in the development of such systems :)

Chris said...

I still think that proper training and repeat testing is a much better solution than trying to stick a plaster (band aid) on the problem.

Paul said...

I thought that lane-departure warnings and blind-spot warnings were dangerous as soon as I heard about them. They will absolutely decrease people's awareness of their surroundings while driving, no question (in my mind).

Auto-wipers - not against the idea but the only time I've driven a car with the feature (a rental) it was more of an annoyance than a help, especially when the rain was very light.

Auto-park - Good for someone who has back problems and has trouble turning in their seat but other than that it's a feature that decreases driver skill.

Auto-lights - I actually like this feature. Does it make me a lazy driver or lull me into a false sense of security? I don't think so personally but that's just me.