Development of eCOMPLEX Software for the Control and Safety Assessment on Crytical Infrastructures.

There is a huge number of critical infrastructures whose failure is related to a high potential risk. The majority of all these critical infrastructures counts with an important quantity of sensors and other measuring equipment controlling a great number of variables. They were installed in order to assess the safety level of the installation.

Data series generated by sensors growth considerably during the operation of the infrastructure. In a short period, databases could store thousands and thousands of records that contains very useful information of installation’s behavior. However, some analysis techniques commonly used are not taking advantage of the knowledge caught in data. So, it could be very beneficial the usage of the most recent techniques of Machine Learning and Data Mining in order to reach a better understanding of the state of safety of the analyzed infrastructure.

The aim of iCOMPLEX project is the development of a computational framework capable of managing the information generated by data sensors on a group of infrastructures. This management comprises a series of tasks: gathering and storing entries into database; display of a diversity of graphs and charts adapted to expert requirements; and, especially, all the processes regarding data analysis, combining a series of interesting pre-processing functions along with configurable machine learning tools.

All these tasks can be useful to make behavior predictions based on data. These predictions can be compared with readings in order check the normal behavior of the installation, and it may be possible to detect the inception of anomalous events that could put the safety under risk and trigger a catastrophe.



This R&D project has been funded by: