A self-driving Audi sports car has reached speeds of 120mph at a race track in America.

The modified Audi TTS, nicknamed 'Shelley' by the Stanford University students who worked on it, managed to make its way around the track at full speeds without a driver behind the wheel.

Instead, the car uses an array of cameras, radars and computer algorithms to see where it is going and figure out the correct speeds to take.
Although the sight of an empty car speeding round a track may be quite a spectable, mechanical engineering professor Chris Gerdes and his team of students spend whole time glued to screens, analysing the car's performance during the lap.

Although the car gets up to 120mph on some stretches, it spends most of the lap driving between 50 and 75mph, the range of speeds at which most car accidents occur.

By contrast, Google's driverless car is limited to only 25mph.

By analysing a huge amount of driving data, the team can get a look at how the robotic car adjusts its throttle, brakes and makes use of the friction in its tires during the race.

Understanding the computer's decisions could help in the development of automatic collision avoidance software in future.

Gerdes said: "A race car driver can use all of a car's functionality to drive fast. We want to access that same functionality to make driving safer."

There's already some similar systems on the nation's roads - the Autopilot feature in the Tesla Model S can automatically keep the car in its lane, make it change lanes, and even emergency brake much faster than a human could.