A multi-viewpoint system to support abductive reasoning
Information Sciences: an International Journal
How to learn from the resilience of Human-Machine Systems?
Engineering Applications of Artificial Intelligence
Using the BCD model for risk analysis: An influence diagram based approach
Engineering Applications of Artificial Intelligence
Hi-index | 0.00 |
The paper discusses on the concept of human errors when operators use barriers or when barriers have to be designed to decrease risks associated with human behaviours. It gives taxonomies of barriers and of their uses and develops the concept of the human-error-based design of barriers to control human errors. Human error assessment methods are then proposed regarding three dimensions: retrospective analysis methods, prospective analysis methods and on-line analysis methods. A new methodology including these dimensions is proposed to take into account several uses of barriers: normal uses, unintentional erroneous uses, intentional diverted uses and uses of new barriers. In order to achieve the identification of these barrier uses, this approach is based on the comparison between the prescriptive and predictive behaviours for a prospective analysis or between prescriptive and real behaviours for a retrospective analysis to define new barriers or to redesign the existing ones. A management of the learning process of such barrier uses is integrated in order to make this design or redesign processes possible.