As a San Francisco Bay Area driver, I see lots of strange stuff on the freeways. A Honda Civic painted to look like a zebra, a station wagon covered entirely in seashells, and an Econoline van painted purple with spiked rims. These are sights you learn to accept, but never quite get used to. And then there was the morning that I saw my first Google Self-Driving Car.
You can’t help but look; whenever there’s a car with a big maybe-camera, maybe-megaphone on top, you just have to read the side. Even here in the heart of Silicon Valley, you never stop looking out for these strange vehicles. I felt it somewhere in my soul – this one was different. I sling-shot around a Mazda 3. I sped past an Audi a5. I had to read the side door. As I approached the Lexus caravan, my heart skipped a beat. The words printed ever so neatly on the driver-side door: GOOGLE SELF-DRIVING CAR.
I knew my dreams were about to come true.
I turned, I looked.
You can imagine my dismay when I spotted a regular, non-robot human being in the driver seat. We made eye contact, and he laughed as he saw my face drop from the smile of a little girl finally getting that pony she always wanted, to someone who realized they owed several months pay in back taxes. No robots. Just a boring guy getting paid to take a chauffeured tour of the 680-S freeway during post-rush hour traffic.
For those of you who haven’t seen one of these things, here’s a reference photo.
It looks like a regular car, but inside it’s equipped with advanced technology that enables it to drive itself. How does that even begin to work? Sadly the answer is not “magic” or “robots” like I wanted it to be – but it is, actually, quite simple.
The self-driving car is loaded with a combination of gyroscopes, accelerometers and altimeters. A gyroscope is a clever little device used for measuring and/or maintaining geographical orientation. When combined with an accelerometer (for measuring acceleration) and an altimeter (used to measure altitude), the internal computer on the vehicle can provide incredibly accurate global positioning, far outreaching the standard GPS found in our current, manually operated models. So rather than asking “was that the exit? Did you see what number that was? Did that sign even have a number?,” the car simply knows where, on the planet, it needs to go, and it goes.
Well, if this was the Australian outback and roads and people weren’t an issue, we could probably just stop there. But we can’t, because while this car DOES have to drive through the outback, it also has to drive through rush hour traffic across the Bay Bridge. And let’s say you were trying to get to Treasure Island – there’s a chance it could think the shortest route would be off the bridge, through the water, and straight on ‘til Alcatraz. And THAT would be BAD.
It needs more than just super-GPS.
So, the self-driving car is also covered in cameras and sensors, and employs the use of radar and linar. The cameras are used to track relative distance of the vehicle from its surroundings; they read traffic lights and street signs, and watch for lane-lines and road edges. This is basically an intensified version of the technology that already exists in some of the partially automated consumer vehicles that are already on the market. You know the ones that don’t run over your kid’s bike, or parallel-park for you because you’re really bad at it? Yeah. Those. The radar, which uses radio waves, and the linar, technology that uses pulses of light, are employed in mapping nearby obstructions, as well as maintaining distance from other vehicles and pedestrians. It’s sort of like how dolphins use echolocation – except that in this case, there are millions of other dolphins all trying to swim along the same current, and tweet about it simultaneously.
And that is exactly why the automobile industry thinks we need self-driving cars like Google’s. Mounted conspicuously on top of the car is a 360-degree, 3D camera; the thing that practically screams “I’M A ROBOT HERE TO TAKE OVER YOUR WORLD” is said to provide stronger situational awareness than ever capable of a human driver. Why’s that? Because no blind spots, and no distractions. The camera doesn’t need coffee. The camera is never worried about a presentation. The camera isn’t trying to get the kids to stop throwing cheerios out the window at the neighboring Mercedes. The camera just watches. It watches, it tells the computer where to go and how fast to get there, and it doesn’t have an Instagram account.
With the growing intensity in the race to perfect these self-sufficient cars comes one glaring question: Why? Why do we need cars that drive themselves? As a society, have we really just gotten THAT lazy? Or that irresponsible? Do we just hate driving? Well, sort of. But that argument doesn’t look as great on paper.
Instead, automotive makers and techono-wizards like Google have posed a few other suggestions for the mass production of self-driving cars. The first, and strongest being the argument of safety. With the introduction of these cars on the freeways, the number of auto-related deaths would most definitely diminish. To err is human, and traffic accidents are a pretty common err. Handing over the reigns to a series of algorithms takes the humanity right out of driving. Will there still be accidents? Well, has your Gmail account ever gone offline for 5 minutes and then come back all wonky-looking? Yes. Yes it has. But 95% of the time, it works. And right now, I pass at least 3 accidents a day during my commute. I’d trade up to those Google-mandated odds.
The other driving force, driving force – do you see what I did there? is the opportunity for bettering the mobility of the elderly and disabled. With a car that can self-brake for pedestrians, and self-drive the speed limit, those who no longer, or have never, qualified for a driver’s license, can once again rely on automobile transportation. With this advancement, those individuals will gain a sense of independence and freedom that was once impossible, and that’s pretty cool.
But the age-old question remains, can you truly trust a computer to brake, or turn, or swerve every time when you need it to? That’s putting a lot of faith in the very science that had us crying in corners thirteen years ago, mumbling Y2K.
These self-driving cars are no Tesla Model S just yet; they have miles to go as far as meeting safety standards and regulations. This is why I was met with a Fry and not a Bender in the driver’s seat of that Google Lexus. The technology is in beta. SUPER beta. Cannot be trusted on an old-dirt road beta. These cars have to be supervised at all times, accompanied by a “backup driver” to slam on the brakes if for some reason Google forgot to disable the energy-saver “put computer to sleep after x number of hours” function.
The benefits of this advancement are exciting, and many in number. Fewer accidents, less fuel consumption, as many of the models in the production stages are electric cars, and also, less time commuting. The existence of these cars will change the very structure of cities and metros, enabling less dependence on needing to live close to jobs, restaurants and bars. Which leads me, of course, to the greatest benefit of all; the only reason I needed to be convinced that this particular technological marvel was essential to the betterment of our world: the self-driving car is the permanent designated driver that pub-crawlers and bar-hoppers can have at the ready.
I mean, just imagine:
Wanna go out and get insanely drunk?
Who’s gonna drive?
Me, too, friend. Me, too.
Eleanor Thibeaux is a writer, audio engineer and cake-enthusiast living in the Northern California Bay Area. She can often be spotted on the freeway in her Scion TC dancing and singing along with Disney show tunes during in bumper-to-bumper traffic. You can’t hear her, but she sounds REALLY good. She can found on Twitter, and on her blog, Grumpy Girl Monologues.