Friday, September 07, 2012


"Brake! Brake now, you mickey mouse piece of....!!!!"
-John Spartan, Demolition Man

Self Driver
The DARPA Challenge is a prize offered to anyone who can come up with a truly safe self-driving car. So far Google appears to be in the lead, having tested their self-driving car for many hours and over 300,000 miles without incident. One thing is certain: unless something major happens to mess up the world, we're looking at a near future with self-driving cars.

At first blush, it sounds incredibly dangerous, letting a computer drive your car. How will it react to emergencies, will it know what to do when various events occur, can it deal with problems with the car? That's all being tested but I am confident that it will be worked out and almost all of those concerns have been very well dealt with already, if not all of them. And yes, a lot of the supposed navigation programs have had serious problems but they are getting a lot better and more reliable.

Obviously at first these are going to be premier luxury items, far beyond the reach of the ordinary car buyer. Just the R&D costs alone are going to push the price up, and the technology is quite high. Doubtless insurance will be costly at least for a time as these cars are first introduced onto the open market. My guess is they won't be big sellers at first, but I expect they will definitely become more popular as the price comes down.

Its true that Americans especially love to drive, so the appeal of a self-driving car isn't as great as it might seem to some. Getting out on the road alone and driving can be quite pleasing to many, and is a way of relaxing or dealing with stress for more than a few people. And the power and freedom of driving yourself is one of the main reasons cars overtook horses so rapidly around the world.

But the truth is, if all you're doing is going from point A to point B, a car that drives its self would be pretty nice. And having that option would allow you to relax and ignore the road if you have to, sort of like switching on autopilot on an airplane.

Imagine taking a trip from Seattle to Denver, and being able to just enjoy the scenery. Imagine a self-driving RV that you can go hit the bathroom or sit down and watch a video, play some cards, read a book or something instead of having to focus on driving. Tired of that commute every day? The self-driving car can do it for you, so you can ignore the road, read your email, text, watch that cat video, or whatever.

Which brings us to a piece by Walter Russel Mead in which he asks "is high speed rail irrelevant?"
Not only are innovations like these relatively cheap to implement, they are much better suited to meeting the needs of a large and sparsely populated country like the United States. It’s becoming clear that the infrastructure lobby, with its love of expansive, expensive rail networks, is fighting an increasingly irrelevant battle.
And he has a point. Instead of building a multi billion dollar (perhaps trillion dollar) infrastructure of rails and trains that history has proved people don't really care for in America, self-driving cars seem to be making the entire exercise irrelevant. If your primary concern is getting from point A to point B but not having to worry about the drive, why take a train that only goes to specific locations, costs extra, and leaves you without a car at the end of the ride?

Instead of high speed rail, private enterprise is coming up with a solution that costs the taxpayers little (should cost nothing, but the prize is out there from the federal government) and is even more convenient and useful. For long distance America, self drive just makes more sense.

But there's a catch. People can hack into your computer, your cell phone, even your On*Star in your car. What if they hack your self driving car? At Popular Mechanics, they bring up this concern, and others. Its already been proven that you can decoy a GPS positioner to think its somewhere it is not. What if a self-drive car gets into a problem with an aggressive driver?
[Mascarenas'] work starts with teaching robots defensive driving techniques. Consider the Precision Immobilization Technique, or PIT, the standard way cops stop a fleeing vehicle (see a good example in the video below). The pursuer draws up and gives the rear end of the target a sideways nudge, causing the target vehicle to slew sideways, lose control, and stop. Mascarenas thought unmanned vehicles would be easy prey for robbers using this sort of tactic to stop trucks and steal their cargo. So he developed techniques for the robots to escape, evade, and recover from the PIT maneuver.
Probably the best option is to not sleep while the car is driving and be ready to take over when need be. But one way or another, this is going to be part of our lives by the end of this decade, I expect.

1 comment:

texas defensive driving said...

After reading the article, I felt like trying an auto-drive car would be so much worth the risk. Taking all the considerations about one’s safety in using this car and the convenience it gives to people dealing with boredom when it comes to driving the car in all too familiar directions, I must say its high price is just right. With the advantages it offers, I wouldn’t be surprised if this goes out in the market around the world in less than a decade.