Times are changing. Technology is improving at an exponential pace. That means that as technology improves, we get better at improving technology. Just 50 years ago, simple computers were so large they required entire floors to store them. And these computers had less computing power than what you would find in a modern low-quality calculator. The smartphones which we use on a daily basis, the devices which have become nearly ubiquitous, would be almost unrecognizable to the individuals who helped create early computers.
It is almost impossible to project what the future is going to look like, but at this point, we know one thing for sure, the future is going to involve self-driving cars. Already, self-driving cars made by Google and Tesla are performing test drives on our streets. Some people believe that the technology is already advanced enough that self-driving cars should be open to consumer purchase. Whether that is true or not, we are still many, many years away from the day when all cars on the road are of the self-driving variety. Before that day comes, you may want to familiarize yourself with how self-driving cars work. Let’s dive into it.
First things first. When you sit down behind the wheel of a self-driving, well, you won’t sit behind a wheel at all. Most prototype self-driving cars do not have steering wheels or gas pedals. Instead, everything is handled internally, by computer.
Google likes to break the self-driving process down into four distinct questions. The car asks itself:
Where am I?
What’s around me?
What will happen next?
What should I do?
First, on the “where am I?” step, the vehicle uses laser range finders, cameras, sensors, and GPS data to determine exactly where it is in the world. It doesn’t just know what city or state it is in. It knows what street it is on. It knows what lane it is in.
Next, the car asks, “what’s around me?” It uses those same sensors, cameras, and laser range finders to create a clear map of its immediate environment. These hi-tech devices can identify everything skyscrapers to pedestrians to bus stops to small animals. They can detect how far away everything is. Have you ever watched the show Daredevil on Netflix? It is something like that.
Now the car must ask, “what will happen next?” You may not be consciously aware of this, but when you are driving you are constantly making small predictions regarding the other vehicles in your environment. You are anticipating what they are going to do. A self-driving car must do this as well. It uses motion detectors, inertial navigation units, and complex predictive probabilistic software to anticipate where objects are going. It can also categorize objects for better accuracy. For example, if the car detects a two-wheeled object traveling at 45 mph, it can assume the object is a motorcycle rather than a bicycle. It will then use that information to enhance its predictive capabilities.
Finally, the car asks, “what should I do?” Software within the car quickly plots a trajectory and begins moving at a safe speed. It can perform complex maneuvers such as giving bikers more room, yielding for pedestrians, or merging onto the highway.
The most impressive thing about self-driving cars? They perform each of these steps THOUSANDS OF TIMES PER SECOND. They react faster than humans and are better at paying attention to the road than humans. On day soon, the road will be a safer place thanks to self-driving cars.