[Zurück]


Vorträge und Posterpräsentationen (mit Tagungsband-Eintrag):

W. Sihn, F. Ansari:
"Industrial Data Science - From Raw Data to Useful Applications";
Vortrag: 24th International Seminar on High Technology, Sao Paulo; 10.10.2019; in: "Digitization of Production and Digitized Production", K. Schützer (Hrg.); (2019), ISSN: 2175-9960; S. 3 - 32.



Kurzfassung englisch:
In the light of rapid technological enhancements and digitalization, complex process-related, technical and organizational interdependencies and correlations in production and logistics can no longer be grasped and resolved by domain experts exclusively. Modern data science methods and technologies, e.g. advanced machine learning, are needed to overcome the increasing complexity, to identify new potentials and ultimately to drive business values. From theoretical perspective, huge amounts of data have to be transformed into concrete conclusions and recommendations for action and decisions. However, in contrast to other business sectors, such as financial services, production industries, especially SMEs, suffer the lack of data quality and availability, resources for extensive data mining processes as well as data science competencies. Therefore, industrial data science projects should necessarily focus on how to generate new data with the help of industrial IoT (IIoT) and how to build simple but usable and accurate data models in cooperation with domain experts e.g. for predictive maintenance, process time forecasting or real-time collision detection for human-robot-collaboration. Furthermore, SMEs gain benefits from new approaches such as text mining or reinforcement learning applied for prescriptive maintenance and advanced production planning and control.
Finally, yet importantly, achieving intelligent functions in the industrial value chain is explored by identifying challenges and potential directions toward the integration of artificial intelligence (AI)and industrial processes.

Schlagworte:
Data Science; Production and Logistics Management; Data Mining; Machine Learning; Artificial Intelligence

Erstellt aus der Publikationsdatenbank der Technischen Universität Wien.