Unlike Tesla, which just presented to the world its “almost 100% self-driving car“, Google has not yet presented to the public their own cars, and definitely will not do it before 2019.
Google is adopting a different strategy: Google is studying ways to make fully autonomous car, without requiring the pilot input. When Google began to deal with self-drivings (in 2010) the plan was to build similar auto-pilot systems to those of Tesla, which would have required so that the pilot intervene at the slightest problem.
However, in 2013 by videos taken from internal cameras, Google realized that the drivers distracted almost immediately, playing with the internal display, looking at the map, looking at the phone etc. Driver were even removing their hands from the steering wheel and even falling asleep, relying completely on autopilot despite the car went to nearly 100 kilometers per hour. Google engineers realized that maybe you can’t expect a driver to be distracted and try to avoid an accident in a few seconds.
Google then decided to take a completely different strategy, designing a system that would exclude the pilot from the equation: no pedals, no steering wheel and a maximum speed of 40 km / h. No car designed for long trips, but for getting around the city.
Completely self-driving cars, according to the Washington Post, will change our lives in the next few years, and for the better, “the elderly and the blind will be able to get around needing much less assistance. The deaths in road accidents will not disappear completely, but will decline. The cars will talk to each other and will drive in a more predictable, so the traffic will decrease. Americans no longer have to spend so much time looking for parking. Commuters will be less stressed. Many choose not to own a car, because it probably will take advantage of the inexpensive steps of car services unmanned. And the country will waste less fuel. “
Google prototypes exploit a laser detection system known as LIDAR, which analyzes the environment around the car to create a detailed map of what surrounds it. Tesla cars instead, exploit radar sensors that make the cars only semi-autonomous: they can stay alone in the middle of a lane or change it, and increase or reduce speed depending on traffic and obstacles. According to a scale designed by the National Highway Traffic Safety Administration (NHTSA), an agency of the US government, to assess what are autonomous cars of this type, Tesla cars are Level 2 self-drivings. Google, for instance, aims to levels 4 or 5, the first provides that the driver enter the path information, and then the car do the rest;
Tesla has always stated that its autopilot system is only meant to help the driver and not to replace it: in literature and on demonstration video explained that the driver must remain vigilant and keep your hands on the steering wheel or at least in close proximity. The Tesla system was presented last October, and Brown was one of the customers who were trying to preview: people who have decided to try the software update Tesla had driven a total of more than 200 million kilometers before ‘Brown accident. It is a way, among other things, to create a relationship of loyalty and trust with customers, and attract “those willing to take risks for a potential technology can change the world.”
The system adopted by Tesla is far more rudimentary and essential than that of Google, and therefore is much less expensive and above is already available on the market. As explained by the Los Angeles Times, “the meteoric approach Tesla is in line with its image as a small but significant automaker that is shaking the industry, while Google – a technology company from which nobody expects car – has the luxury of time ‘: Tesla Motors makes cars, Google does many things including, for some years, even cars.