Car pooling or paid ride sharing should be computerized, at least for commuters. Here's how a good system might work, using currently available technology. It could even be the next great application for computers, wi-fi, and the internet.
Way back in the 1960's or 70's I read that almost any mass transit system looked good compared to people commuting one to a car, but if you could average four riders per car, almost nothing imaginable was as efficient, at least for fuel and probably for capital outlay as well. After all, the cars are already bought, the owners already know how to drive them, and the current road system is more than adequate if commuting traffic were cut by three fourths. The problem is making four to a car attractive enough, especially convenient enough. Car pooling gets to be a drag very quickly if anyone needs to change his schedule.
Truckers, even solo independent truckers, are using the internet to match loads to trucks that need to go someplace anyway. Web sites apparently post loads available with departure points and destinations so a trucker who just dropped a load away from home can find another load to haul on his trip back home. In stead of phoning around, sometimes for days, a trucker can go online and then telephone to confirm his return trip load in half an hour. That's great turnaround time for a trip that can take several days. But a typical commute takes 30 to 90 minutes, so thirty minutes wait to fill up your car, or to find a full car ready to go, is a gross inconvenience. To be competitive, car pooling almost needs to be several minutes faster than walking from your parking place to your desk at work, just to make up for the time lost while you watch one or two other guys being picked up in the morning or dropped off at night. What's more, it still has to be flexible, so if you want to leave early or late, you can still get a ride nearly as quickly. Looking up web sites is too cumbersome for this. Commuters need some sort of very rapid computerized matching of riders, empty seats, departure points, and destinations.
What if you could just telephone the 'commuter computer' in the morning when you were ready to leave, and tap in a few extra numbers to confirm who and where you are and your destination. That could be preprogrammed into your home and mobile phones. If a car came to your house in just a few minutes, and only stopped to pick up one other person, and dropped you off right at the door to work after dropping off only one other person, it might actually be better than driving yourself on a day with bad weather. That could be done if everyone who wanted to drive had a laptop computer in his car, or a car computer, that had mobile internet access. As the driver started to work in the morning, he would turn on his car computer and notify the commuter matching service of his destination and the number of empty seats he has. The car computer should have G.P.S., so it can report where the car really is. Within seconds, the matching service should give the car computer the information on the best route to pick up riders, and they should all be nearby. The G.P.S. navigation in the car computer should help the driver find his riders -- they may vary from day to day and he may not know the way to their houses. The car computer should then figure the best route to work, dropping the passengers off so the driver ends up near his own destination.
On the way home, you might 'swipe' something like a credit card with your name and home address into some sort of card reader located near a good pick up point. The card reader could add its location and send the ride request to the ride matching computer. Within minutes, a ride should come to the pickup point for you. Commuter cars might even have some sort of display mounted on the rearview mirror, facing outward, that would have your name in flashing lights, so you'd know the right car. If you ended your work day away from where you started, you wouldn't need to go back to get your car. Businesses where you might want to stop on your way home could give you cards with their address on the 'swipe strip,' and with an ad and their business hours, to boot. ("Had a tough day? Stop at Louie's Bar on the way home! After all, you're not driving!")
The passengers should expect to pay for their ride, and they should also expect safety monitoring, in the form of a computer camera that would send a picture of the inside of the car to some central 'filing' computer every few seconds. Those files could be discarded when they were a week old, but the pictures should deter any crime from happening in the commuter cars. Public transportation doesn't allow for much privacy; at least commuting in a monitored car should be safer than riding a bus or subway. After all, it's hard to pick somebody's pocket when he's sitting down, and it's dangerous to try with the camera on. (Just be sure to turn the camera off when you go parking on Lover's Lane.)
Such a scheme might even help put the right number of cars on the road every morning. If the commuter matching program can't find anybody nearby to give you a ride, it could suggest that you drive yourself, and pick up some folks on your way. You should make enough to pay for your gas and wear on the car, and maybe even enough for pricey, convenient parking if you work in a congested area. If you don't own a car, a ride-matching program should be almost as convenient as commuting by taxicab, and a lot cheaper. To avoid stealing too much business from taxis, the program could limit commuter cars to one round trip a day; that should at least blunt any political opposition to the scheme from taxicab companies. Prices for rides should be determined by a formula in the computer program, but the formula could charge more on days nobody wants to drive. It might make getting up an hour early to shovel snow off your driveway worthwhile if everybody's fare went up five bucks. The passengers ought to be happy to pay that much for the privilege of not shoveling their own driveways. Young car owners would have another way to use their cars to help earn the car payment, in case delivering pizza doesn't bring in enough. On ozone action days, the city could subsidize rides, if the computer could write checks or credit bank accounts.
There's money to be made if such a system worked well. If Bill Gates or Steve Jobs could get, say, Houston to adopt a system, and it reduced congestion and smog, the world would beat a path to his door. Every car would need a computer with wireless internet access that worked across a whole city -- even in tunnels and 'urban canyons'. Every car computer would need a commuter matching program and maybe a camera and a lighted display. The camera and display lights might be on opposite sides of the same gadget; it might hang on the rearview mirror. The central matching computer might need to be some humongous supercomputer, but maybe not. It might work better to disperse the matching among all the car computers. That way you might even get to arrange always to pick up your girlfriend, or never to pick up your flatulent brother in law.
A fast enough computer system could basically create an efficient market in rides to work and back, and perhaps in rides to wherever you want to go within a city. It might put some public transportation out of business, but if buses on fixed routes can't compete with custom-matched rides, let's punt the buses. Most of them seem to need subsidies, anyway. Up to now, you couldn't look out your window at the traffic and flag down the car that is going your way. Maybe soon, you'll do just that.