Let the hive mind of Engadget get that for you.
"We need a digital camera that can be switched on and fire off that first shot fast. It's not a commonly tracked statistic on any review site, and nobody seems to have this information for every camera. We were hoping other readers could inform us as to what small digital cameras can fire off their first pics in under a second (ideally under half a second). It needs to be small, but mostly, just really quick in operation. Thanks!"
My freshman year in engineering school, I had to attend a doctoral thesis presentation by a grad-student that was doing his research on this very subject. This was back in 1996 and he was piggy-backing on over 30 years of research at that point.
Long story short, this is a very attainable goal. Over 10 years ago, they already had 8 miles of test road somewhere out in California (too lazy to Google it) and had succesful demonstrations of a convoy of 10 cars traveling at 120mph, spaced only 3' apart.
The jist of this system is simple - a special set of lanes (seperated from the normal lanes) would be put into place. You enter your where you want to get out of these lanes (now, I assume just enter your destination in your GPS). When you enter the special lanes, the car takes over. It merges you into traffic and does your driving, follow embeded sensors in the road, radar, vision and communication with the other cars around you. When a hazard is detected, the vehicle will avoid it and also transmit this information to all the vehicles following it. Basically, the convoy moves like a flock of birds or school of fish.
Once you reach your destination, the car pulls out of the lane you take over again. They even thought of sleeping (or dead) drivers - as you approach your off-ramp, the car requires a manual input. If you don't respond in time it merges you out of traffic and alerts authorites that something is wrong.
Again, I really simplified the system.
And even back in 1996, the equipment and technology existed. There were two main hurdles that had to be overcome - 1) Sell price and 2) Infrastructure. Research has shown that the most people will pay for an "option" on a vehicle, even as great as this, is no more than $3,000. I believe the system they had at that time was somewhere in the $10,000 range. As for infrastructure, it's not really clear who would pay for the advanced traffic control system that would have to be put in place, as well as the construction costs to embed the sensors.
Of course, the best way would be to devolp a system that didn't need either of these and with the advances in machine vision and detection technologies, I believe this will be the way GM goes.
So, this is not as far fetched as it seems. Of course, what everyone else says about the regulatory nature of the US is true, so that will be the biggest hurdle.