Watching Teenage Drivers with Webcams

Over two hundred teenagers in southern Maryland are now driving around with a webcam on the rear-view mirror of their cars. Whenever they turn or brake sharply, the resulting g-force triggers the camera to record a 20-second sequence of what went on inside and outside the car before and after the incident. These dynamic snippets go via the web to a company in San Diego that reviews them, attaches little helpful comments about how such dangerous driving incidents can be avoided, and notifies the teenager's parents that the video is now available for viewing.

Although deaths and injuries in automobile accidents have been declining slowly for years, over 40,000 people died in highway-related accidents in 2007. Anything that makes that number go down without severely compromising some other desirable outcome of automobile use is worth considering. And at first glance, the specter of Mom or Pop looking over the teen driver's shoulder, so to speak, seems like a good idea. A similar study done in 2006 showed that drivers who started out triggering the webcam a lot with their jerky, high-risk driving, eventually learned to reduce their triggering rate (and thus drive more safely) by four-fifths. It's too early to tell whether a similar improvement will result from the Maryland experiment. But one thing is already clear: the teens don't like the idea, even though some grudgingly admit that the system has improved their driving.

Do the teens have a point? Is the webcam an intrusion on their privacy? Obviously it is, but then you have to ask whether the chance of saving someone's life is worth a little less privacy. And it's not like the thing was on all the time. Teens do other things with, and in, cars that I'm sure they wouldn't want their parents to see. But when the company that operates the system says it won't forward anything that's "embarrassing to the teen" in their words, that seems to be enough to satisfy most young drivers. Of course, if the company were ever to betray that trust, the entire system might suffer a black eye that it might never recover from.

This system is just one example of how technology is making it possible to monitor more and more aspects of our daily lives, in ways that were unthinkable back in the days when George Orwell wrote 1984. One of the creepier images of that novel was the spy cameras everywhere, monitored by secret police whose presence the citizens were reminded of through the slogan "Big Brother is watching you." A sure-fire argument against that kind of thing ever happening in reality used to be that you'd never be able to man every camera everywhere, because eventually you'd end up hiring one half of the populace to watch the other half. But notice that the in-car webcam uses smart technology—namely, accelerometers—to select only those incidents worthy of study, thereby reducing the work of human editors to manageable proportions.

So as time goes on, it will be more and more practical to acquire webcam data on all sorts of activities, and still be able to handle the massive amounts of raw input intelligently. Is this a fundamental threat to privacy, liberty, and all that? Or is it a tempest in a teapot?

The answer hinges on those who are doing the spying, or monitoring, or whatever you want to call it. In the case we're discussing today, a private company is involved with consenting families, and if the company does anything out of line, they are liable to lose business fast. That's one of the best constraints against misbehavior. Governments do not have such a negative incentive, which is why government-sponsored monitoring of behavior can be more problematic. A case in point is the increasingly obtrusive safety inspections for airline passengers. In certain airports, systems are now in place that use millimeter-wave sensors to see through a person's clothes. The people who inspect these images are not co-located at the inspection point, but still, you wonder if and when this kind of thing will be abused.

It seems the best thing to do in these cases is to ask whether the system is doing any good. In the case of the in-car webcams, it looks like they may well improve driver safety, which is good for everybody. In other situations, such as in-flight security, it's harder to evaluate effectiveness except with tests in which people try to sneak by the inspection stations on purpose. And the news regularly carries reports that inspectors often fail these tests. On the other hand, we haven't had any U. S. planes get bombed or turned into flying missiles since Sept. 9, 2001, so something is working, at any rate.

The other factor to consider is the continuing decline in monitoring technology cost. The current webcam system costs $900 plus a $30 monthly fee, but if it proves popular, these costs could go down to where it would be offered as an option when you buy a new car. If insurance companies like it, you could get a discount on your teenager's insurance rate if you agreed to install the device. And once it's in there, it will work for everybody: the teens, Mom, Pop, and Grandpa. So one day we may all be driving around with spycams in our rear-view mirrors, who knows? Let's just hope the people operating the cameras then are as trustworthy as they are today.

Sources: An early report on the in-car webcam can be found at the Washington Post website for Oct. 24, 2008 at http://www.washingtonpost.com/wp-dyn/content/article/2008/10/23/AR2008102303821.html. The 2007 automotive fatality statistics can be found at the U. S. National Highway Traffic Safety Administration website http://www-nrd.nhtsa.dot.gov/Pubs/811017.PDF.