Data and information form the basis for the engineering of tomorrow and for modern value creation. In order to lay the foundation for this, interdisciplinary research is elementary - starting with the necessary domain knowledge regarding data acquisition, through the transfer and fusion of the data, to the processing of the information gained from it into knowledge and action. In all these steps, the reliability of the data sources, the human-in-the-loop, and the processing algorithms must be taken into account.

The research institute RIVK specializes in such holistic solutions to gain knowledge from data. The aim is to make the data of modern sensor technology available to the manufacturing and operating industries in such a way that they can process the information without hurdles and thus generate knowledge. This interdisciplinary approach requires engineering knowledge about the design and risk assessment of components and machines, IT and data science knowledge, domain knowledge regarding data sources, and knowledge regarding reliability considerations.

RIVK specializes in the following four research areas:

1. Value Creation/Engineering
   Development of value creation concepts from data. Building on existing probabilistic, predictive, and prescriptive approaches, holistic approaches are pursued that incorporate a wide variety of data sources.
2. Data Science
   Development of approaches and frameworks for fusion and transformation of data and associated metadata into information in the Internet of Things through Semantic Interoperability as well as extraction of knowledge from information through Digital Twins, Artificial Intelligence, and Quantum Computing.
3. Domain knowledge sensors
   Meaningful use of data is only given if the expert knowledge for data science and value creation is given in conjunction with the domain knowledge regarding the data sources. This is the only way to extract information from data. This is especially true for complex sensor technologies such as non-destructive testing, where a deep understanding of the method is required for further use of the data. In many cases, existing knowledge about signal origination is not sufficient for the new value chains.
4. Reliability
   Any data source has an inherent accuracy. When using the data source for value creation, it is essential to know the reliability, quantify it, and make it available to data scientists along with the actual data and metadata. In addition to physical, mathematical, and statistical influencing factors, human factors must also be considered.

Established: February 2020
Directors & Founders: Dr. Johannes Vrana & Dr. Daniel Kanzler