Google Cars have been a secret project on the streets of Mountain View California but Google have gone public with the project and announced it has done over 190,000 accident free miles without human intervention. They use robotic Toyota Priuses and have been driving them in busy city centres and on mountainous terrain. The technology is currently far from being commercially viable but offers an insightful view of where the technology could lead us. The aim of the technology is to reduce road accidents worldwide and ultimately save lives.
How the car works: The car has laser mounted system on its roof. This allows it to generate a 3D map of the surroundings and input that into the cars computer systems. The car combines this laser system with detailed world maps it is able to avoid obstacles, respect traffic laws and,in theory, drive to any well mapped location in the world. The vehicle has four cameras mounted around it for a widespread view and built in GPS to keep the location up-to-date.
The following video goes into more detail on how the technology actually works:
Larry Page and Sergey Brin believe that smarter vehicles could help make transportation safer and more efficient: Cars would drive closer to each other, making better use of the 80 percent to 90 percent of empty space on roads, and also form speedy convoys on freeways.
A really exciting prospect of self driving cars is that the idea of car ownership may become the thing of the past. There is a possibility that individuals will pay a rental fee on driver-less cars, contact them via their smartphone and will have a car arrive at their destination to pick them up. This may however put Taxi drivers out of business. There are many legal issues which affect the future of the autonomous vehicle like who is responsible if the car crashes. With time, one would expect that these difficulties would be overcome but at the moment this and the reliability issues of the cars are preventing commercial viability.
At the current cutting edge of this technology Google has developed an algorithm with the ability to learn. They have set it up to play games of Atari, with only a basic algorithm and video input from the game and the computer was able to learn how to improve at the game to a position where it is better than a human. The technology takes pixel input and can learn and adapt to this information, a technological advance that could have serious implications for autonomous vehicles. If it is possible to race a car in a video game then there is no major reason why it won’t be possible to race a real car. There is potential for the car to be able to drive without pre-loaded maps, the car would be sufficiently smart to see the road in front and adapt accordingly.
The future of driver-less cars may not be as far away as we think but will require several major changes to happen from legal to market pressure from existing traditional car makers. It is however another example of where AI is making a big impact on where we’re going in the next 10 to 20 years.