Applicability of human reliability assessment methods to human–computer interfaces

The UK Office of Nuclear Regulation (ONR) has undertaken a Generic Design Assessment of two nuclear power station designs for prospective construction in the UK. This assessment included a review of the Human Reliability Assessments (HRAs) submitted as part of the probabilistic safety assessments (P...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Cognition, technology & work technology & work, 2013-02, Vol.15 (1), p.19-27
Hauptverfasser: Hickling, E. M., Bowie, J. E.
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:The UK Office of Nuclear Regulation (ONR) has undertaken a Generic Design Assessment of two nuclear power station designs for prospective construction in the UK. This assessment included a review of the Human Reliability Assessments (HRAs) submitted as part of the probabilistic safety assessments (PSAs). Both reactor designs have human–system interfaces driven by digital technology. However, the data and methods for assessing human error probability (HEP) pre-date such technology. Therefore, the ONR sought to establish whether existing HRA methods remain applicable to modern human–interface interactions and hence continue to provide a credible insight into the risk contribution from human error. An extensive literature review was undertaken to identify or derive relevant HEPs. Data have ranged from those associated with particular interface objects, plant start-ups and post-fault diagnoses. There appear to be some interesting paradoxes within the data explored in this paper. Based upon the data reviewed, it is concluded that existing human reliability assessment methods are likely to be optimistic in their estimates of HEPs where diagnosis is involved or where process control is dependent on human–computer interaction. Shortfalls in the availability of published relevant data and the scope of existing HRA methods have been identified by this work.
ISSN:1435-5558
1435-5566
DOI:10.1007/s10111-012-0215-x