Predicting design-induced error in the cockpit

Neville A. Stanton, Don Harris, Paul M. Salmon, Jason Demagalski, Andrew Marshall, Thomas Waldmann, Sidney Dekker, Mark S. Young

Research output: Contribution to journalArticlepeer-review

19 Citations (Scopus)

Abstract

This paper describes the Human Error Identification (HEI) Technique called the Human Error Template (HET). HET has been developed specifically for the aerospace industry in response to Certification Specification (CS) 25.1302. In particular, it is intended as an aid for the early identification of design-induced errors, and as a formal method to demonstrate the inclusion of human factors issues in the design and certification process of aircraft flight-decks, including supplemental type certification. The template-based approach was chosen because it appeared to be quick to learn and easy to use. HET uses a hierarchical task analysis as its starting point. A checklist of twelve (12) external error modes is used to determine which might lead to credible errors for each task step. For each credible error a description is given and the outcome described. If the likelihood of the error and the consequences are both high then that task step is rated as a 'Fail'. The error mode taxonomy developed comprises: fail to execute a task, task execution incomplete, in the wrong direction, wrong task executed, task repeated, on the wrong interface element, too early, too late, too much, too little, misread information, and other. HET was then compared to SHERPA, HAZOP and HEIST. Thirty seven (37) analysts were employed in this study based on a landing scenario. HET showed significantly better Sensitivity Index scores than any of the other methods, and the greatest number of correct error predictions (hits). The results from the HET validation study demonstrate that HET meets all the criteria set. It is easy to learn, the error taxonomy has been specifically designed for flight-deck tasks, it is auditable, and it has been proved to be both reliable and valid. HET is recommended for use in the design, evaluation and certification of aircraft flight-decks.

Original languageEnglish
Pages (from-to)1-10
Number of pages10
JournalJournal of Aeronautics Astronautics and Aviation
Volume42
Issue number1
Publication statusPublished - Mar 2010
Externally publishedYes

Keywords

  • Certification
  • Cockpit design
  • Human error identification
  • Validation

ASJC Scopus subject areas

  • Aerospace Engineering
  • Space and Planetary Science

Fingerprint

Dive into the research topics of 'Predicting design-induced error in the cockpit'. Together they form a unique fingerprint.

Cite this