Robots are awesome. Robotic cars are even more so. I've been following the various iterations of the autonomous car for a while now and I swear they just keep getting better. Which is a very good thing, beyond the pure technical achievement of solving a highly complex problem, because as the population ages, a lot of folks are driving who really shouldn't be.
One of the peculiarities of the U.S. is our reliance on cars; we've got too large a landmass and too scattered a populace for mass transit of the same kind that's common to Europe, say, at least outside cities and towns where population density makes it a more viable option. So, we've got a landscape and infrastructure that's set up for cars and that's unlikely to change any time soon. Where do the driverless cars come in? Well, wouldn't it be nice if older people who really should not be driving themselves still had a car-based option to get around, so that loss of their car keys didn't also mean a loss of independence? We're trying hard to keep people living in their homes for as long as possible and transportation is a critical part of that package. Unfortunately, it's not always easy to separate a driver from their car keys, even when it's clearly a poor idea to have them behind the wheel.
Tom Vanderbilt's excellent article in Wired, "Let the Robot Drive: The Autonomous Car of the Future Is Here" not only does a great job of listing the various assistive technologies that are already creeping into our cars (providing lane departure warnings, collision avoidance, blind-spot monitoring, driver alertness, and parking assistance), as well as talking about the current state of autonomous cars and the great strides that have been made in getting to the point where a suitably instrumented car is capable of safely driving itself in a normal driving environment (including a variety of exciting weather conditions) and populated with the usual mix of non-autonomous vehicles, pedestrians, cyclists, and random wildlife. On the one hand, it's hard to read about these projects and not get really excited because of the integral wow factor. It may not be a personal jetpack, but I'd much rather have a driverless car! On the other hand, we're not there yet—sensors can still be confused, how people should drive and how they actually drive differ, and some of the devices used to enable fully autonomous navigation are expensive—the Velodyne LIDAR used on Google's self-driving car, for instance, ain't cheap.
Autonomous cars promise improved safety and that's a very big deal, since driving is, for most of us, the most dangerous thing we do. They also have the potential to change the way our roads look and function. To get an idea of some of those potential changes, I'd advise reading Emily Badger's article in The Atlantic, "What Intersections Would Look Like in a World of Driverless Cars." Realistically, even when the technical challenges are solved, that still leaves a raft of legal and behavioral hurdles to overcome. Regardless, the next decade or two is going to be rather interesting.