BMW and Artificial intelligence 11.11.2020: BMW #NEXTGen 2020 The basis for automated driving and natural interaction . The BMW Group currently employs artificial intelligence (or “AI” for short) in more than 400 applications and in every relevant area of the company. Development is o

BMW #NEXTGen 2020



The basis for automated driving and natural interaction .



The BMW Group currently employs artificial intelligence (or “AI” for short) in more than 400 applications and in every relevant area of the company. Development is one example; here AI provides the basis for automated driving and the most natural possible in-vehicle user experience.

“Artificial intelligence plays a central role at the BMW Group,” says Simon Euringer, BMW Group, Head of Intelligent Personal Assistant. “It helps us with the processing and interpretation of large quantities of data – whether that’s in sales, production or research.”





Extensive data pool provides the basis for all AI applications.



Artificial intelligence requires an extensive pool of data, whatever the application. In order to virtually reproduce driving situations at the Driving Simulation Centre and, in the next step, train BMW Group vehicles for automated driving, the actual road network and traffic situations first need to be digitalised.



The basis for data-driven development therefore consists of the data collected worldwide by the vehicles of the BMW Group Automated Driving Test Fleet, plus – since the end of 2019 – the data from vehicles owned by BMW Group customers, who have consented to this anonymised information being transmitted and processed.



As of October 2020, this already equates to over 250 million kilometres of real-world driving experience from the regions of Europe and North America. From the totality of data from these two sources – the BMW fleet and BMW customer vehicles – particularly relevant driving scenarios and environmental factors are selected and their relevance is continuously increased.



Particularly challenging scenarios for AI-based simulations are, in turn, extracted from this to ensure that the intricacy of reality is taken into account as fully as possible in the development of complex driver assistance systems like the Driving Assistant Professional and future automated driving applications.



These vast quantities of data are processed by the BMW Group High Performance D3 platform with over 230 petabytes of storage capacity (in the planned expansion) and an extremely powerful computer platform with more than 100,000 processor cores and over 200 GPUs (Graphics Processing Units).




The development of automated driving with the BMW iX.



AI creates an algorithm based on the data collected, and this then calculates safe and anticipatory automated driving functions. So without AI, automated driving cannot progress. All of the results of the calculations are tested directly at the BMW Group Autonomous Driving Campus in Unterschleißheim, which is connected directly to the BMW Group High Performance D3 platform located just a few kilometres away by high-performance fibre-optic cables.



In the future, the BMW iX will play an important role in the ongoing development of automated driving. “With its all-new technology toolkit, its computing power, high-performance sensor technology, highly advanced data-driven development and embedded AI for processing complex tasks, the BMW iX offers the potential to continue developing automated driving functions year after year,” says the BMW Group’s André Roskopf, an expert in data-driven development and AI.



AI can already be experienced in BMW vehicles.



Artificial intelligence can already be used and experienced by drivers and passengers in BMW Group vehicles. The Intelligent Personal Assistant (IPA) – available in a large number of models since the end of 2018 – employs AI to make it easier for customers to use functions in their vehicle.



 The driver activates the IPA using the prompt “Hey BMW” and can operate a host of functions by voice control without having to use pre-defined commands. “The key here is AI and machine learning. These are indispensable elements of natural interaction – i.e. intuitive vehicle operation – and enhance both the comfort and safety of those on board,” explains Simon Euringer.








Matthias Schepke- photos BMW