Estimated 1.3 million people die from car accidents every year. That's equivalent to about ten Hiroshima bombs going off each year or two international airplanes crashing daily.

With such staggering statistics, some may argue that people legislatively should not be allowed to drive. Well, that's obviously not a viable solution. Our society has unanimously decided that the benefits automobiles bring far outweigh the drawbacks of it. Thus industries have been diligently seeking to find a pragmatic answer to address this issue.

Vision Zero, goal of reaching zero collisions and zero emissions, is the ambition that ZF Group and numerous other companies in the industry carry. No one, however, has been able to figure out how to fully prevent accidents in a case of human error—such as drunk, distracted, and drowsy driving—that accounts for 94 percent of all accidents. The world with no collisions, therefore, had been deemed rather absurd and unobtainable for decades.

That was until autonomous vehicles were introduced—idea of taking humans out of the equation.

Cars of the future may be born fully autonomous, no driver needed or wanted.

Mechanism of a driverless car

It's important to first examine the main components and functionalities of driverless cars to see if they will be safe enough on the roads with no human surveillance. By and large, sensors and computer hardware and software will be incorporated into the system to give the vehicle its "eyes", "ears", and "brain".

Sensors

Driverless technology requires various sensors that detect objects around the vehicle and bring data from the car's exterior. There are digital cameras that function as the "human eyes" and capture visual surroundings of the car in a stream of real-time data. There is also Lidar (Light Detection and Ranging) that provides a three-dimensional, 360 degree digital model of the environment, called point cloud.

Radar sensors, using electromagnetic wave echo system with transmitters and receivers, can detect the position and speed of an object 170 meters away. Ultrasonic sensors, or sonars, use sound waves to see close range tiny objects at especially high resolution. There are also other detecting devices, such as the high-definition GPS and the IMU, that are accountable to provide additional data to the vehicle's operating system.

Artificial Intelligence

Then come the control systems that interpret the data provided and decide what the car physically does. Autonomous vehicles use data-driven AI, machine learning and deep learning techniques. Using complex neural network systems, driverless cars are able to learn to drive with increasing amount of data.

An analogy of this connection would be the relationship between fuel and engine in a spacecraft. If the engine is massive but has insufficient fuel, the spacecraft wouldn't go very far. On the other hand, if there is plenty of fuel but has a tiny engine, the spacecraft wouldn't be able to fly far off the land. Similarly, deep learning algorithms operate as an engine, requiring strong structures and foundation of the system, and data act as fuel, the driving force behind the system.

With advanced GPUs and hundreds of microprocessors that operate the software and the data from sensors, autonomous vehicle's control system continues to improve exponentially. This technology can "see" and "think" far better, further, and more efficient than any humans can.

Driverless world

Aside from these top-notch technologies, what would a driverless world look like?

1. The very first difference to be noticed would be the silence. There won't be people honking at each other for switching lanes too abruptly. Sirens and horns won't be necessary in a society with no human drivers. We would experience a much quieter world.
2. With just a button on a phone, a car would be brought right in someone's driveway, and the cars will drop them off at specific locations.
3. Parking areas would be better utilized, as cars will be able to park themselves. Parking spaces represent 57 percent of the total downtown space in Houston, Texas. Roads could be built more narrow with cars that can calculate their horizon and projectiles significantly better than humans.
4. People would be able to work, sleep, or watch TVs during their commutes. On average, people in the U.S. spend about 30 minutes commuting each way, an hour a day. Commuters would have a choice to make best use of that time.
5. The number of accidents would be remarkably reduced, which is by far the greatest driving force behind the advancement and development of this industry. With more and more enhancement to such technology, reaching a world with zero collisions is a possibility.
6. As discussed in the beginning, emission rates will drop conspicuously. Not only can driverless cars calculate and drive more efficiently, but they will also have to accommodate for what's called engine lag time. In a standard gasoline-powered engine, the lag time between converting chemical energy into mechanical energy is notoriously unpredictable. Thus it is challenging to start the car with precise timing. The best solution for this issue is to exchange the standard engines to electrical engines, encouraging industries to invest heavily on the development of electric vehicles.

Road ahead

Not many know that this technology that could save millions of lives and revolutionize the transportation system will be available in the near future. Both software and automotive companies have been working diligently on driverless cars, forming partnerships and investing millions of dollars in the programs. In fact, most top-tier companies have committed on releasing fully autonomous vehicles to the market by 2025, and some say earlier. In less than a decade from now, cars will be driving themselves everyday on regular roads.

The finish line for Vision Zero is in a foreseeable future. Just need to endure and finish strong.