Data analysis methods in automatics and manufacturing

This area covers very wide applications of data analysis in the issues of estimation, classification and prediction. Specific areas of application include, but are not limited to, process and device diagnostics, fault prediction, load prediction, phenomena prediction or condition-based maintenance. The methods used in the research are mainly based on Bayesian statistics and selected areas of machine learning.

In this field, work is carried out on data processing, construction of classifiers and detection of events in time series.

Manufacturing control and management systems


In this field, research is conducted in two main areas. The first is the development of ERP software in the field of data science. Extending the functionality of such systems with analytical and predictive tools gives the possibilities of wide applications and significant benefits. The second area is the optimal control of the production process. This issue is related to the problem of optimal control. We work on the development of effective computational methods to optimize real manufacturing problems that may include unexpected disturbances and non-deterministic faults.

Numerical methods in system control

Our teamwork on effective numerical methods addressed to the problem of implementing methods and control algorithms and signal processing on various hardware platforms. The issues of approximation and implementation in real-time of filters and regulators with characteristics of incomplete order systems are considered. We develop algorithms and design methods to provide increased control system robustness.