What's That Black Box Doing Under My Hood?

IN ANN ARBOR, MICH., THERE IS A family with a Ford F150 pickup truck. In the last two months they've taken it on 465 trips, for a total of 2,000 miles. They've moved the gearshift lever 1,200 times, sped down a highway at 87 miles per hour and let the truck idle each morning for approximately 15 seconds. How do we know? Because every few weeks during this wholly unremarkable stint of driving, engineers at Ford's research center dialed up a cellular modem in the truck's engine and downloaded the data, everything from engine temperature to how often its owner had hit the brakes. In the last year Ford has put these dial-up devices in 350 vehicles around the world. It's the first lap in a race that could change the way we drive.

It's been 100 years since the first cars were mass-produced in America, yet scientists know more about the mating habits of obscure African insects than they do about the driving habits of Americans. About all today's researchers know is what we tell them, and self-reported surveys are often inaccurate. Now that's changing. Inside new cars are tiny sensors and computers that track the engine's every gurgle and grunt. It's simple enough to build a system that saves the data; the trick is harnessing it. From Detroit to Washington, researchers are looking to the car's internal diagnostic computer as a sort of Nielsen system measuring how we drive. The goals: more durable vehicles, safer roads and cleaner air.

Right now engineers design cars based largely on guesswork. They've long assumed, for example, that drivers frequently gun the engine and don't let cars warm up. Wrong on both counts, says Brian Mahoney, the executive in charge of Ford's data-collection project. In the last year his researchers have also learned that some elderly drivers go just as fast as young whippersnappers, that Boston drivers really are dreadful and that police cars idle constantly (no word if doughnut stops are to blame). They've also come across oddities like the lady who made 40 thirty-second trips one afternoon (turned out she was changing parking spaces at an outlet mall). Ford engineers are already planning tweaks to design vehicles better suited to these kinds of real-world conditions.

While tougher cars would be nice, safer ones would be a blessing. Crash investigators would love the kind of data that come from ""black boxes'' after airplane crashes--car-accident analysis today is imprecise, involving measures of skid marks and dents to estimate speed and force. Most airbag sensors record a car's motions just before a crash, but thus far the data haven't been collected. ""We have a real absence of data on how severe crashes are in the real world,'' says one federal safety official. ""We probably should have been doing this from day one with airbag-equipped cars.''

Better data on where and how we drive would also be a godsend for people who study highway patterns and mass-transit systems. Researcher Peter Gordon of the University of Southern California says current information on how Americans commute ignores how often we run errands on the way to and from the office. Those chores are one reason people don't car-pool or use mass transit. If experts understood real driving patterns, they could design more efficient commuting methods, or at least understand why people choose to live where they do. Better data would also let the EPA calculate cars' emissions more accurately, says University of California economist Charles Lave, since today's emissions ratings use just a rough model of typical driving habits.

The big hurdle to all of these projects is privacy. Few drivers want Big Brother watching them roll through stop signs, and no one wants his car's electronic diary used against him in an accident lawsuit. Ford pays folks to let it capture their cars' signals; safety experts hope to gather data by promising anonymity. But in a world where every credit-card purchase leaves a trail, perhaps driving will become just another virtual footprint.