INTRODUCTION Digitization is changing the world. Technology, innovations and especially ever-growing data ecosystems are turning the industrial economy on its head. Markets have long began to adapt to the changing environment and are remodeling business concepts and rethinking value chains. Data is the new oil, and an intelligent exploitation of data within industrial processes and products has become a key success factor for generating new revenue and optimizing the way value is created. Manufacturing companies, for example vehicle manufacturers, have a keen interest in data that provides insights about their products, in this case vehicle usage, functionalities of digital assistants or quality. Such data can provide knowledge, that is valuable for developing new services like autonomous driving. Customer interaction can be improved too, for example by being able to address drivers in particular situations, such as when the first indicators for serious wear appear. Collecting and intelligently analyzing data forms the fundament for all these activities. In this context, more and more so-called digital twins are being used, i.e. digital copies of physical objects. These may be vehicle components or the vehicle as a whole. In vehicle development digital twins offer particular potential for optimization, for they enable developments to be largely tested on purely digital simulations. This, in turn, offers major potential savings in terms of time and money. Up to now, vehicle parts and control software have been developed, then tested on many thousands of kilometers of test tracks in order to get results that can trigger improvements. This process takes several months and, depending on the outcome of the testing, has to be repeated more than once until the required vehicle properties and quality requirements areachieved. Now digital twins can simulate vehicle components in their real-life contexts, which means they no longer need to be physically produced or tested on test tracks. Rather, existing test data can be used on digitally simulated components and thereafter be evaluated. As a result, engineers get results far more quickly, but at the same time they face challenges linked to the use of new, digital technologies such as big data analysis and predictions by artificial intelligence. The main challenges are the provisioning of suitable hardware and sensors, the efficient management and processing of gigantic volumes of data and the development of intelligent analysis algorithms. In this way data can be converted into insights and then again into design improvements. The prerequisites for this new type of engineering have been put in place by the technical innovations of the recent years, for example the ubiquitous networking, the availability of less expensive hardware and sensors, and the constantly improving capabilities of data storage and data processing. This white paper is aimed at engineers, developers and IT managers who also face digital engineering challenges. It provides a clear answer to the question of how big data technologies can be used to improve development processes in the control software area. 4 5