Audi is using a model car to develop its self-parking technologies, which actually looks like a lot of fun.
Each year, the Conference and Workshop on Neural Information Processing Systems (NIPS) presents advances in the fields of machine learning and computational neuroscience, and this year, Audi is showcasing its expertise at the conference for the very first time. The German automaker is demonstrating how an Audi Q2 model car aids in developing intelligent parking strategies by showing off the 1:8-scale model car and its smart parking process.
According to Audi, the model car autonomously searches for and finds a suitable parking space in the form of a metal frame, inside an area measuring 3×3 meters. It can then park itself there with no human interaction.
The model car consists of two mono cameras, facing forward and toward the rear, along with 10 ultrasonic sensors positioned at points all around the car. A central onboard computer converts the data into control signals for steering and the electric motor.
The system also learns through trial and error, starting by selecting its direction of travel at random. From there, an algorithm autonomously identifies what actions were successful, helping refine the parking strategy. Audi claims in the end, the system is able to solve even difficult problems on its own.