Audi is showcasing its expertise on how a car develops intelligent parking strategies with the help of a scale model at the Neural Information Processing Systems (NIPS) conference in Barcelona.
Audi revealed that self-learning systems is key for automated driving cars. This is portrayed by their 1:8 scale model car, called the ‘Audi Q2 deep learning concept’, which demonstrates an intelligent parking process. On an area that measures 3x3 meters, it autonomously searches and finds an appropriate parking space before parking itself there.
This is possible due to the Q2 deep learning concept’s sensor technology which comprises of two cameras, facing forward and towards the rear, along with ten ultrasonic sensors positioned at points all around the model. A central on-board computer converts their data into control signals for steering and the electric motor.
The model car’s parking ability is made possible by deep reinforcement learning which essentially means the system learns through trial and error. To begin with, the car selects its direction of travel at random. An algorithm autonomously identifies the successful actions, thus continually refining the parking strategy. So in the end the system is able to solve even difficult problems autonomously.
Audi is expected to use the deep learning-based software for the first time in 2017 on the next generation Audi A8 which will enable piloted driving in congested traffic situations as well as piloted parking.