HOW TO IMPROVE ERROR ANALYSIS AND CLASSIFICATION SYSTEM

An existing system for human error analysis and classification is discussed. The lack of a comprehensive model of human error for this system decreases the quality of human error investigation and, as a result, the quality of recommendations that addressed safety-related deficiencies. The model suggested in the present paper is based on theory of frames, modal logic and psychological experiments and permitted the construction of meaningful description for error-provocative situation. The database for such situations gives the opportunity to find the most frequent causes, and therefore to mitigate the deficiencies. On the basis of the described model, the new decision support system for analysis and classification of error is proposed. This system provides both the data collecting for human error investigation and analyzing the most frequent causes through the database. The suggested system demonstrated a higher performance level than the existing classification system in analyzing the most frequent causes of error, and could enhance either accident investigating or incident reporting systems in aviation