Due to advancing digitalization and the Internet of Production, sensors collect ever-greater amounts of data. This data must be examined and analyzed in detail to make best use of it. While statistical and stochastic approaches as well as machine learning are suitable spotting aberrations or making diagnoses, most human observers find it rather cumbersome and feel that understanding the results requires great effort.
Selection, design and preparation of data
For this reason, not only data and results of the analysis but also connections, conclusions, prognosis and underlying processes, too, have to be prepared and presented clearly. Suitable methods for data visualization and software dashboards can present relevant data clearly and render interaction simpler and more intuitive.
Methods of data visualization
The Fraunhofer IPT not only works on choosing formats of visualization but also designing intuitive software interfaces for data and knowledge management. We also investigate how to prepare results of the analysis as suitable management summaries to simplify the decision-making process. In this, we aim to neither present data in too complex a manner, not to leave out or even distort decisive information.
- Selecting visualization methods
- Designing intuitive software interfaces
- Preparing results as management summaries