Do you remember learning how to parallel park? How many times did you hit the curb before you got it right? After lots of practice and learning from your mistakes, you became a pro at it.
The same is true for Audi's new driverless car. This vehicle can learn from its mistakes to become better and better at parking itself.
The German car company recently demonstrated this technology with a car model at the Neural Information Processing Systems conference, which was held in Barcelona, Spain earlier this month. The model, Audi Q2 Deep Learning concept, can look for and locate a parking spot and then maneuver itself in between the lines.
It uses a front-camera, a rear-camera, and several sensors that allow the miniature vehicle to navigate itself. The information from the camera and sensors is transmitted to the car's central computer, which in turn controls the brake and accelerator.
Artificial intelligence is a major component to fine-tuning this technology. Most computers are programmed to carry out certain functions but machines with A.I. can use trial and error to improve. The model uses an algorithm to determine successful movements in order to create a better parking strategy.
Society may not be riding around in autonomous vehicles yet but we seem to be getting closer and closer. We can also expect a self-driving car from Audi in 2017. It will be an A8 Sedan that can drive as fast as 37 mph and complete some parking functions. Google, Tesla, GM, and so many other manufacturers are also perfecting the autonomous vehicle technology and running test drives.
More Car Tech Stories