Saturday, December 29, 2012

What's So Great About Self-Driving Cars?

Self-driving cars.  Oh, man, the holy grail of the modern-day commuter.

As someone who survived a year of long-distance romance with my now beloved wife, and as someone who used to spend up to two hours a day in the car commuting, and who now spends at least an hour a day in the car commuting, I have to say that I've dreamed about nothing more (except the aforementioned lovely) than I have about the self-driving, or autonomous, car.

During may of these hours, I spent time fantasizing about something I felt was inevitable.  That someday, the car would drive me wherever I wanted to go.

I still do this, by the way.  Fantasize about the future of the driving car, I mean.

And over the past few years, I have been following the news of Google's self-driving cars with fascination.  I am so excited that they are making great progress.  I have even taken the trouble to write to my state representatives to urge them to make Illinois a trial ground for the cars.

Here's how I think the self-driving car can help us:

No more distracted driving
Life is busy.  We have fewer hours in the day to accomplish more.  We have alerts coming at us at all times.  To keep up with all of it, lots of people, right or wrong, have taken to driving distracted.  Especially when they face the dreaded Chicago commute, with so much time spent in the car, people feel the need to cram in a meal, read or respond to email, talk on the phone either for home or business, or even get a shave or put on makeup. All that leads to needless fatalities and injuries due to people not paying attention.

The self-driving car changes a lot of that.  If the car is able to take you fully to your destination, then you can read the paper, get in your meeting, eat your breakfast, do some work on a laptop.  Basically, anything you could do on the train.  The self-driving car basically becomes your personal train, going exactly where you need it.

Fewer Accidents
Remember, a self-driving car doesn't have blind spots.  It never forgets to check them.  Their reaction time is much faster than human reaction time.  I don't envision cars, no matter how they are piloted, ever being able to completely eliminate accidents; the situations that can occur on the road are far too varied.  But with tireless sensors and alert algorithms, it's a lot better than what a human being can bring to the table.

No more wasted time
While not as good as telecommuting, the self-driving car means that you can do other things during the commute.  You will be able to organize your day around having some tasks that you are able to do while commuting.   If you are needed for 8 hours of work a day, there may be no reason you can't do some of that work while commuting.  Editing documents, working on spreadsheets, sending email, communicating to clients, writing code, all those things become possible to do while on the commute.

Can you imagine how your life changes when instead of leaving your house at 7 a.m. to be at work by 8 a.m. and leaving work at 5  p.m. to get home by 6 p.m., you get to start working when you get in your car at 8 a.m. and stop working when it pulls into your driveway at 5 p.m.?  Two extra hours at home with your family is a prize we should all be working diligently toward.

No more vehicle ownership
Ok, one of the best benefits of self-driving cars is my favorite.  Tell me why, if cars can drive themselves, would I ever want to have one parked at my house?  Why would I want to have the extra square footage in a big drafty space attached to my home that is just there to house two or more vehicles while they sit idle?  Totally wasteful.  I can pay less for a house if I don't need all that extra space.  Houses that currently have garages can be retrofit to have a new great room for entertainment or can trade that extra space in for storage so they can refurb their basement.

Imagine having a standing order for a vehicle of a certain size/luxury with a service organization whose car comes and picks you up at a certain time every morning.  You jump in and it takes you to work and drops you off at the door.  You want a luxury car?  You pay a little more.  You fine with a little car that gets great mileage and has a low total cost of ownership?  You get a discount.  You pay the automobile bill at the end of every month, and only pay for what you use.

You won't have to worry about owning a vehicle - all that mess of when do you buy, when do you sell, depreciation, repairs, etc.  All those costs are spread out over the car company's vehicle fleet and included in the per-use price.

While you're at work, the car can be part of a fleet making more trips picking people up and dropping them off.  If a single car can fulfill the needs of three or four people through a single day, then there only need to be about a fourth of the cars in existence (note that because people still need to go where they go when they go now, that doesn't necessarily reduce the number of cars on the road at any given time, just the total number of cars needed overall).

No more parking lots
If a car drops you at the door of your work, your office doesn't need a parking garage to hold cars.  Or maybe it doesn't need such a large lot.  Certainly the cars have to be somewhere, but if they are out providing services to other passengers, fewer cars will sit parked all day.  This means that there's lots of useful land that can be converted into something more useful, or just into lush green campus for the businesses that used to need so many parking spaces.

No more drunk driving
I don't know about everyone, but I suspect the only reason people who drive drunk is that they don't want to wake up and have to go fetch their car.  If you don't own a car, the question is moot.  The car comes around and picks you up just as it would do if you were sober.  It's a similar situation to not wasting time, but I separate it out here because it's a very specific problem that this solution absolutely does away with and in a way that doesn't shame the person in any way.

Fewer moving violations and traffic stops
And with self-driving cars, what's the likelihood that there will be traffic infractions?  If the car is following its programming, it knows how fast to go, what the traffic laws are, and won't do things it's not supposed to do.  There won't be any speeding tickets. Check out this article on what it truly means to be self-driving, and how one gives a ticket to a self-driving car.

Better commute times
What the heck causes the commute times to suck so bad?  I am not a civil engineer, but I have my guesses.  Merging always sucks.  Any time three lanes become two, or two become one, or an acceleration lane merges into moving traffic, there's a little dance between the lanes.  I've seen people refuse to let someone in; I've seen people try to graciously let each other in, only to end up in a you first, no you first game.  Doesn't happen with self-driving cars (assuming they all are.  I can see people being jerks and specifically not letting a self-driving car merge), as they have no ego about who goes first.

Computers don't get apprehensive about traffic density.  When traffic density gets too high, people tend to hit the brakes in apprehension.  Self-driving cars behave consistently and would self regulate.  They might move more slowly with higher density as safety dictates, but they wouldn't be likely to set up a standing compression wave in traffic that can last for hours.  Check out this link for some amateur traffic analysis.

Autonomous cars mean no more gapers delays or sun delays.  The car doesn't need to slow down so the person can get a good look at whatever's on the side of the road (and remember, there would be fewer reasons for anyone to even be on the side of the road), and the car doesn't care where in the sky the sun is, if at all.

Ok, so I'm totally convinced.  Now what?
Where do we go from here?  For me, it's a foregone conclusion that this would be one of the greatest advancements our civilization can make, and I want to see this happen here in the USA at the fastest rate possible.

Thing is, I would love to be part of this solution, but I don't know how to help.  One obvious solution would be to try to get on the project at Google.  I'm a technologist with a lot of experience, so maybe one of the things I could do is try to go to work there.  That's one angle, but I believe they probably have all the engineers they need on this project.

I see one of the greater obstacles to adoption is going to be FUD in the public eye.  Despite the record of human drivers vs. the record of self-driving cars, there is always going to be worry that the programmers forgot something, some rare situation that may crop up.  That in that situation, the car won't know what to do the way a human would.  You hear about fatal bugs that crop up in software from time to time, and no matter how remote or infrequent the possibility, and no matter what the empirical data eventually shows, some people are always going to believe they are better than a machine.

Further, there's some buzz that a car must always have a "driver", and that the "driver" must not be drunk.  That kind of outmoded thinking comes in direct conflict to the Utopian vision I have for self-driving cars.

So how can I be part of the solution?  If you are working on these problems, or know anyone who is, I would love to be in touch with you or them.  You can get in touch with me through contact information found at  Even if it's to have a quick conversation about what it's going to take to make this happen over the coming years... What can I do to help?

No comments:

Post a Comment