Abstract
On petroleum facilities, safety equipment is installed to obtain risk reduction. To ensure that the desired risk reduction is achieved during the operational phase, follow-up of the reliability of such equipment is important. Essential means to follow up safety-critical equipment are the reporting, classification and analysis of failure data. However, experience shows that considerable manual effort is needed to attain high-quality data, and therefore, this report investigates possibilities for automated failure reporting and failure class determination. The proposed approaches involve the potential to guide the user with pre-defined help-texts for improved failure reporting, the possibility to utilize relevant information from systems such as the safety and automation system etc. for automatic determination of failure mode and detection method, the potential to automate failure class determination based on standardized failure mode and detection method hierarchies, as well as the use of technical language processing (TLP) methods to extract key information for improved and efficient failure class determination.
