InTrans / Aug 18, 2014

Driving into the future: Self-driving cars

Go! Magazine

Self driving car
This image provided by Google shows a very early version of Google’s prototype self-driving car. The two-seater won’t be sold publicly, but Google on Tuesday, May 27, 2014 said it hopes by this time next year, 100 prototypes will be on public roads. (AP Photo/Google)

posted on August 18, 2014

Although self-driving cars are still years away from being put on the market, the technology for them exists today. Automated features are not new. Many of the vehicles being produced today actually possess very basic forms of self-driving technology, although you may not have realized it. Did you know that the anti-lock brake system as well as a vehicles’ cruise control functions are among the simplest forms of self-driving technology? The most complex systems currently available are:

  • Lane-keeping system:It uses mostly cameras to detect lane lines. The driver is alerted if they start to cross one.
  • The pre-safe system:It uses radar to scan the outside environment. If the system detects that a collision may occur, it prepares for the vehicle to be quickly stopped as well as prepares deployment of the airbags. In some cases, the system can even stop the vehicle itself.
  • Self-parking vehicles:Although currently restricted to only a handful of models, self-parking vehicles use radar, lasers, and cameras to scan the immediate area. An internal computer collects the information, calculates the most efficient way to park, and executes the maneuvers required. Currently in development is a vehicle that lets the driver drop themselves off at the curb before it parks itself. When they’re ready, an app on their phone calls the vehicle and picks them up. Audi is one of the companies working on this type of self-parking car. However, the project is still in the testing stage and won’t be available to the public for several years.
  • Sensors and cameras:The most common type is rear-view cameras, which alert the driver to objects that are in the vehicle’s path while it moves backwards. More complex versions also feature side-sensors that can alert the driver to the presence of a vehicle that may be in a blind spot.

What about fully self-driving cars?

Could this be the future? In addition to the sensors and cameras on self-parking vehicles, self-driving vehicles have a hodgepodge of additional equipment that acquires millions of bits of information every second. All of the equipment is connected to an Artificial Intelligence (AI) system that processes all of the data and, well, doesn’t need a driver.

Google’s self-driving car
Google’s self-driving car. Courtesy of Flickr user Hydrogen.

Self-driving vehicles are outfitted with:

  • Lidar: A set of lasers that rapidly rotate 360 degrees, constantly scanning the surroundings. The computer creates a 3D image based on the scan and uses it to determine if there are any potential obstacles.
  • Radar: Technology that is used to determine the speed of objects around the vehicle.
  • Cameras: Cameras read road signs and detect traffic light changes, which the computer interprets and reacts to accordingly. Cameras can also work with the radar to determine what an object is and if the vehicle needs to respond to it. They can even be used to avoid potholes.
  • Gyroscopes: These devices determine the tilt of the vehicle. When performing a sharp turn, the computer can change the power and breaking levels of each wheel to keep the car from flipping.
  • Accelerometers: Their job is to detect the speed of the vehicle. While the camera is “reading”? road signs, accelerometers can read and follow posted speed limits. The computer can change the speed of the car to remain at or under the limit.
  • Artificial Intelligence: No, it can’t talk to you. However, it can process the millions and millions of data collected every second and use it to make multiple decisions in a fraction of the time it would take a human to process and react to a single piece of information.
    Can cars talk to each other?

In other words:yes, they can! It’s called the Vehicle to Vehicle (V2V) communication system. All vehicles with this system to communicate their intentions with nearby vehicles, allowing them to react to each other. If one car slows down, it alerts those behind it. The system also collects information on the condition of roadways, which is constantly updated and shared with other vehicles. This feature allows vehicles to drive places they haven’t been but still have updated information about road construction, detours, or the location of potholes. The company that’s furthest along in self-driving car development is, surprisingly, Google. Their fleet of self-driving cars has logged 700,000 miles with only two collisions, neither of which were the fault of the system. Most major auto companies are looking into self-driving vehicles, though not all are in development. Those that are, like BMW, are working on duel systems that allow for both self-driving and manual driving.


There is still debate about whether self-driving cars should even exist at all. Let’s look at the pros and cons.


  • Decreases the numbers of collisions
  • Always obeys posted signs (i.e. They come to a full stop at stop signs rather than slowing then rolling through:otherwise known as a “rolling stop.”?)
  • More efficient parking
  • People who are otherwise impaired are able to drive


  • Doesn’t work well in snow or in areas with obscured/missing markings and signs
  • Confusing liability (i.e. Who is at fault in a collision? Is it the owner, the programmer, or the person who got hit?)
  • System bugs and malfunctions
  • Cyber-security
  • Decrease in experienced drivers and increased dependency on technology
  • Lack of laws (i.e. As self-driving cars become more common, laws will need to be made to regulate them.)

By Brandon Hallmark, Go! Staff Writer

Go! Magazine Article Index