The main goal of DAYSTREAM project is the development of data-driven applications for early recognition, real-time tracking and optimal prediction of mobility events that are safety- and process critical. For this purpose, new data ressources will be created through the automatic interpretation, extension and semantic interlinking of existing data. Parts of these data will be made publicly available through the mCloud, which is a mobility data portal of the Federal Ministry of Traffic and Digital Infrastructure. Together, new and existing data form a basis for a variety of future applications. As part of the project, a variety of structured and unstructured data sources are continuously analyzed, evaluated and (if necessary) interlinked for real-time event detection. Due to the massive amount of data considered in the project, the analysis of data streams is done by applying state-of-the-art Big Data technologies. For the real-time data processing, Apache Flink ( a scalable, highly efficient stream processing platform) is used. The main goals of the DAYSTREAM research project are the implementation of massivly parallel procedures of Machine Learning, time series analyses, clustering and anomaly detection. The DFKI is coordinator of this joint research project. The partners are DB Sicherheit GmbH, idalab GmbH, University of Kassel (IfV-FG Verkehrstechnik und Transportlogistik) and Rhein-Main-Verkehrsbund Servicgesellschaft mbH. DAYSTREAM is part of the “mFund” research programmeand is funded by the Federal Ministry of Transport and Digital Infrastructure under grant agreement no. 19F2031A.
DB Sicherheit GmbH, idalab GmbH, Universität Kassel und die Rhein-Main-Verkehrsverbund Servicegesellschaft mbH.