TSimulus: generating time series
In addition to helping our partners to understand these new technologies and methodologies, we develop tools that meet their different needs in this area. For example, in order to test data processing solutions, we propose TSimulus, a realistic ‘time series generator’. The TSimulus project provides tools for specifying the shape of a time series (general patterns, cycles, importance of the added noise, etc.) and for converting this specification into time series values.
Here are three examples of the implementation of our know-how in this field:
AGC has analytical tools to determine the quality of the glass produced in their plants. CETIC has contributed to organizing these tools in the cloud by centralizing the collection of data from their sensors and adapting their algorithms to distribute the computing load and substantially reduce the analysis time.
As part of this research project, we collected the messages exchanged by the components of the Belgian rail network (signal systems, switches, etc.). Analysis of the historical data of these communications allowed the detection of divergent processes within the considered components, as well as the prediction of failures of some of these components. Infrabel was able to offer predictive maintenance of its facilities.
Cloud computing of the data generated by the sensors installed into the Prayon facilities aims to facilitate the understanding of their processes. Consequently, prescriptive analyses can assist production managers by proposing a better configuration of the systems for which they are responsible. Other types of analysis highlight the causes of declines in quality and performance in the different phases of production.
Contact us for more info about our activities in this field.