The aim of this project is to develop fast and reliable tools for the reliability and risk analysis of, and Real-Time Fault location (RTFL), in complex systems as they are encountered in industry. These tools should be easy to use and economical thus encouraging SMEs to perform reliability analyses on their products.
IRAS is a Unix-based software package that constitutes a new approach to the understanding, monitoring and real-time fault diagnosis of industrial process operation. It provides a user friendly graphical environment for modelling processes in a generic way that could provide considerable advantage in generally accepted reliability assessment methods such as: Failure Mode Effect and Criticality Analysis (FMECA) and Fault Tree Analysis. Since Cause-Consequence Diagrams and Fault Trees can be build automatically. IRAS is capable to perform a Real-Time Fault Location (RTFL) action whenever a serious process deviation is observed.
To date Fault Time Tree Analysis, FMEA (Failure Mode Effect Analysis) and RFTL studies have mostly been considered independent. All these algorithms require knowledge about fault imitation within the system and fault propagation through the system. Computerized path development could speed up reliability and risk analyses substantially, but although several attempts have been made, their benefits have been limited by inadequate modelling. Some shortcomings of these modelling procedures are that :
-models described fault propagation in one direction only which made the models dependent on the system variable being analyzed
-they complicate the automatic detection of control loops which is essential for automatic analysis.
When the modelling strategy is available these studies can be integrated into two reliability analysis systems :
-a consistent and compact modelling procedure suitable for computer-aided components and system model development
-an expert system (IRAS : Integrated Reliability Analysis System) which allows the convenient and rapid creation of compact component and system models and is able to perform FMEA and RTFL virtually automatically.