Context analysis to improve algorithms in condition monitoring

Most of the papers describing the use of algorithms for condition monitoring, a data-based discipline that detects and analyses the condition of machines, do not take context into account as a determining factor for the implementation of these methodologies. That is to say, they overlook factors that influence machine monitoring.

The thesis “On the use of context information for an improved application of data-based algorithms in condition monitoring” published by Kerman López de Calle, a researcher working for the Smart Information Systems Unit at the Tekniker technology centre, a member of Basque Research and Technology Alliance (BRTA), has the main objective of analysing the role played by context in terms of machine condition monitoring in three different situations and focuses on the constraints and resources that exist in each one of them.

More specifically, the following application environments are analysed: variable operation wind turbine gearboxes; rotating machines with unknown degradation evolution although stable operation; and a diagnosed electromagnetic drive assisted by a physical model.

Furthermore, the context of each case is studied to propose solutions via algorithms that have been developed to improve machine condition monitoring. The conclusions reached can be used for the three scenarios that have been analysed and to address other monitoring problems.

This thesis has been supervised by Basilio Sierra Araujo, a professor of the Department of Computational Science and Artificial Intelligence at the University of the Basque Country (UPV/EHU) and Susana Ferreiro del Río, a researcher working for Tekniker’s Smart Information Systems Unit.