• DocumentCode
    3449652
  • Title

    Human error prediction in man-machine system using classification scheme of human erroneous actions

  • Author

    Kohda, Takehisa ; Nojiri, Yoshihiko ; Inoue, Koichi

  • Author_Institution
    Dept. of Aeronaut. & Astron., Kyoto Univ., Japan
  • fYear
    1997
  • fDate
    29 Sep-1 Oct 1997
  • Firstpage
    314
  • Lastpage
    319
  • Abstract
    While the advanced technology improves the reliability of technological systems, the reliability of human operators still remains limited. From the viewpoint of systems safety and reliability, the possibility of human error occurrence must be reduced. At the design stage of a man-machine system, it is important to predict what kind of interactions will lead to a serious accident. This paper presents a prediction method using the classification scheme of human erroneous actions, which shows the general result obtained in such fields as psychology and cognitive engineering. The human behavior cannot be described in the same way as the technological system behavior. The context dependency is one of the characteristics of human behaviors. The context determines the most probable human, erroneous action and specifies man-machine interactions in the system. A simple illustrative example shows the details and merits of the proposed method
  • Keywords
    forecasting theory; human factors; man-machine systems; pattern classification; reliability; safety; human behavior; human error prediction; human operators; man-machine system; pattern classification; reliability; safety; Accidents; Educational institutions; Humans; Information processing; Man machine systems; Marine technology; Prediction methods; Psychology; Safety; Space technology;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Robot and Human Communication, 1997. RO-MAN '97. Proceedings., 6th IEEE International Workshop on
  • Conference_Location
    Sendai
  • Print_ISBN
    0-7803-4076-0
  • Type

    conf

  • DOI
    10.1109/ROMAN.1997.647002
  • Filename
    647002