Tesla Model S recently launched its highly talked-about and eagerly anticipated feature, the Autopilot.
It is rather evident that an autonomous, self-driving vehicle requires technology, which is backed by enormous amounts of information for the car to function safely, while driving on public streets. The more data it gathers, the more efficient and risk-free its driving behavior can become.
No wonder then that Musk and his company are forming highly detailed road maps via GPS. Tesla Model S vehicles are all connected to the company using the cloud technology, but there are about 60,000 Tesla models currently equipped with the necessary cameras and sensors for gathering data from the road. Musk calls it a "fleet learning network", meaning he's creating a huge database with the help of all contributing cars – the company processes data and feeds it back to all the cars. "When one car learns something, the whole fleet learns it," said Musk, emphasizing that this kind of 'learning' is what makes Tesla's mapping of the Earth different to all other known self-driving systems today.
The current autopilot feature doesn't allow full autonomy just yet. Instead, drivers are able to choose between various self-driving features, like auto steering, lane changing, self-parking and side-collision avoidance.
The electric cars have not yet reached that phase, when they would be responsible for any occurring accidents. The car is still entirely every individual driver's responsibility, so it is highly advisable to keep both hands on the steering wheel at all times; however Musk expects his electric fleet to go fully autonomous in about three years.