Researchers at Stanford University have created an autonomous vehicle that has lapped a racing circuit faster than a professional driver.
The team of scientists used an Audi TTS, named 'Shelley', which they had programmed to tackle the Thunderhill Raceway Park in California.
The car negotiated the track at speed of 120mph, and recorded consistent lap times.
Compared to racing driver and CEO of Thunderhill, David Vodden, the autonomous Audi completed the circuit nearly half a second faster.
The team achieved the impressive result by studying drivers' brain activity to understand the neural circuits at work when negotiating a difficult racetrack, and implementing that into the development of the vehicle.
They are predicting that with development, the technology could endow cars with skill of iconic racing drivers such as Michael Schumacher.
"Race car drivers are really fantastic using all the friction between the tyre and the road to get around the track," said programme director Professor Chris Gerdes to The Telegraph.
"Now they are doing that to be fast but the same mathematics holds whether you're a race car driver trying to go around the corner without going off the track or spinning, or whether you're a normal driver going on an icy road where you come in a turn to fast and you want to stay in your lane.
"We've got the point of being fairly comparable to an expert driver in terms of our ability to drive around the track."
The boffins at Stanford aren't alone on using autonomous driving tech to increase performance. Last year Audi sent a self-driving version of its flagship RS7 saloon around the Hockenheim F1 circuit.