Audi are one of the many automakers currently embroiled intensively in driverless car research. They believe the next step will be machine learning-powered driver assist features that can function in tight situations such as driving in traffic and parking. The new features are planned to be part of the next generation Audi A8 saloons, to be released in 2017. Before the car hits the salesrooms however, the technology has to be put through its paces; ensuring its infallibility by the time it reaches the public.
The German car maker has found a novel way to do this. The features are being tested on a much more compact model than the planned A8 – more accurately, a model eight times smaller than the real thing.
Watch the tiny vehicle in action in the video below. Using deep reinforcement learning (machine learning that operates on a trial and error basis with rewards given for the computer when it gets things correct), Audi’s technical whizzes have managed to program the model so it can locate and park in a 3 metre by 3 metre space.
“An algorithm autonomously identifies the successful actions, thus continually refining the parking strategy,” writes Audi in a blog post. “So in the end the system is able to solve even difficult problems autonomously.”
Several other manufacturers in the car industry are also using machine learning to develop autonomous driving (startup Drive.ai is developing a similar program) but Audi is probably the biggest, and technology it introduces will be further reaching.
Get your Audi service with Servicing Stop – the UK’s most convenient car service provider with FREE collection and delivery. Book now for anytime in the next six months!