Audi Piloted Driving Uses “Deep Learning” Technology

Audi’s new driverless technology goes so deep you never know where you’re going to come back up!

June 04, 2015: HERNDON, Virginia – With Audi trying to make its piloted driving impeccable, its engineers are depending on an advanced technology that is referred to as “deep learning” by developers. The technology is used for training computers to resemble the processes of human brains.

The “deep learning” technology is being used by Audi in order to speed up the development of self-reliant driving systems.

Audi has collaborated with Nvidia, the renowned chip manufacturer, in order to develop an “automobile-computer model” which resembles the way in which new information is processed by the human brain.

The autonomous Audi A7, referred to as ‘Jack,’ was used to demonstrate the capability of the artificial intelligence algorithms, which are programmed into the zFAS driver-assistance control that is as small in size as a tablet. It is being used by the A7 as a brain for the management of the independent driving system.

According to an official statement from the company “The Audi processor … analyzes every frame of video that comes in, and it senses edges which it groups into shapes…It learns that the shapes are objects, then learns to differentiate those objects.”

The cameras in the system gather the visual information which is interpreted afterwards by a computer and logged then into a database. This data is then used to refine the technology and develop it even further, which can improve the recognition of the objects as the four-wheeler clocks more distance.

The entire procedure includes massive amount of data measuring in terabytes and has to be processed within the shortest possible time. Instead of transporting the data to the cloud, the system depends on an Nvidia Tegra processor integrated into the vehicle.

According to Audi, the improvements in artificial learning will assist the autonomous technology to progress fast to the production level from the stage of prototype.