12-11-2018 05:56 PM
Thomas Kaelberer, Manager of Intelligent Interior, Daimler
This week at the Xilinx Developer Forum keynote in Frankfurt, Ivo Bolsens, Xilinx CTO, was joined on stage by Thomas Kaelberer, Manager of Intelligent Interior at Daimler. During the exchange, Daimler revealed for the first time additional details around its technical collaboration with Xilinx first revealed in June of this year.
At that time, Daimler and Xilinx disclosed that the two companies were collaborating on an in-car system using Xilinx technology for artificial intelligence (AI) processing in automotive applications. Kaelberer was now able to disclose that this collaboration is contributing to a first-of-its-kind AI solution in the new Mercedes GLE Sport Utility Vehicle.
Photo courtesy of Daimler
The system, called the MBUX Interior Assistant, recognizes the occupants’ natural movements and their context so the vehicle can predict driver and passenger needs and accommodate them without the need to push any button. Daimler showed some of the impressive capabilities including a ‘searchlight’ function that can understand when you want to see inside the car at night and the MBUX Interior Assistant will respond by automatically turning the lights on and off with just the extension of the driver’s arm.
While it is well known that AI solutions are being designed to drive cars autonomously, Daimler has also aimed to change the experience inside of the car by bringing complex deep learning algorithms to a car’s interior. According to Kaelberer, the trend toward greater driving assistance features combined with longer times spent in traffic has driven the need for a higher level of automation even inside the vehicle. Instead of always pushing buttons, now Daimler offers an interior that intuitively interprets how the car can support you in given situations.
When designing the MBUX Interior Assistant, the Daimler team was faced with a unique challenge. It had to design the entire MBUX Interior Assistant computing subsystem within the roof of the vehicle, which is a very thermally constrained environment, particularly in the summer. Daimler engineers benchmarked many powerful computing platforms, but Xilinx was chosen because it had the best performance-per-watt and lowest latency, combined with automotive grade qualification. The teams from the Daimler development centers from Germany, India, and the U.S., in collaboration with Xilinx, successfully completed the ambitious project in less than two years. This was accomplished in part thanks to the enabling software provided in the Xilinx AI platform.