Kurzfassung

 

 

The "Socio-Physical Advanced Research Cloud Infrastructure" (SPARCI) is a DFG-co-funded large-scale equipment (DFG: Großgerät) that provides a powerful computer cluster with a connected long-range wide-area network infrastructure (LoRaWAN infrastructure) for researchers and lecturers. The high-performance infrastructure is the basis for extensive research projects involving the mining, storage, analysis, and use of large amounts of data (Big Data) in sociotechnical...


 

 

The "Socio-Physical Advanced Research Cloud Infrastructure" (SPARCI) is a DFG-co-funded large-scale equipment (DFG: Großgerät) that provides a powerful computer cluster with a connected long-range wide-area network infrastructure (LoRaWAN infrastructure) for researchers and lecturers. The high-performance infrastructure is the basis for extensive research projects involving the mining, storage, analysis, and use of large amounts of data (Big Data) in sociotechnical (human-machine) and cyber-physical (mechanical and electronic things) systems. Data sources and data recipients are derived from research on information systems in enterprises, public administration, and the public World Wide Web. The project pushes into new areas of intelligent linking of real and physical objects with their virtual environment and their use in the aforementioned domains.

 You can find the guide to the cloud here.

The envisioned research has very high computational power and storage capacity requirements for exploring growing datasets and executing artificial intelligence and machine learning methods. In particular, the development of deep learning algorithms and neural networks are leading to a new need for GPU-based computing power that can reduce the computation time required to process existing and growing datasets to a usable level and which is provided by the equipment.

 

To collect sensor data, the large-scale device includes a LoRaWAN infrastructure that will be both stationary on the university campus and usable in off-campus research projects. This infrastructure enables wireless, energy-efficient, bidirectional transmission of, for example, sensor data over distances of several kilometers and also allows geolocation of transmitters using multilateration and time-difference-of-arrival methods via the arrival times of data packets at multiple receiving devices.

 

The research groups involved have extensive prior experience in machine learning, innovative approaches to Web Science, collaboration systems, Internet of Things and IT concepts in public administration. The specific combination of specialisations of the involved research groups enables the necessary interdisciplinary, holistic approach. To maintain compliance aspects, the project will develop a concept for storing large amounts of heterogeneous sensor data, taking into account aspects related to data protection, data security, and privacy of individuals.
» weiterlesen» einklappen