Published:
2018, TRANSPORT MEANS 2018 - PROCEEDINGS OF THE 22nd INTERNATIONAL SCIENTIFIC CONFERENCE, Kaunas, Kaunas University of Technology), p. 709-713), ISSN 1822-296X
Annotation:
The paper focuses on the issues concerning enhancement of the safety data management through understanding of the internal data structure and system behaviour. To build a strong system foundation based on safety knowledge, and to find a system on how to gather such knowledge are the key issues in modern safety engineering. Systematic data collection is not a new concept, however, the focus was always on quantity, while data quality was often misunderstood or hard to understand and be dealt with. A research topic was therefore focused on quality and content of the available data, covering the state of the incoming data, their wider classification, analytics and utilization through instruments creating safety intelligence. Having a large amount of data, means having an ability for potential system behavioural pattern identification. This is however strongly influenced by data structure and applied analytical methods. Identifying behavioural pattern in other words means articulating a certain signal generated through system, carrying information regarding system state, functioning and potential deviations. Due to a fact that such system manifests as a stochastic one, it is highly influenced by certain level of internal or external noise. The goal of the research is to examine system abilities for behavioural deviations detection and elimination of the negative, unnecessary and inadequate structures in data gathering process.