Human Error Classification System And Data Collection Method Based On Socio-technical Approach

In 1990, the known psychologist James Reason of Britain Manchester university put forward the concept of "latent failures", which have caused the extensive concern of the scholars of various countries immediately. Reason think any mistakes and breaking rules or regulations are only essential but not sufficient conditions of accident under defense-in-depth, they are only triggers of accident, only when these triggers and the ability of defence-in-depth happen to meet, will the accident be just caused. Unfortunately, after accident happens, people often only notice the manifestly active failures and ignore the latent failures which hide most deeply and threaten safety biggest.Latent failures are interpreted as organization error and technological error, Using the principle of the cognitive system engineering, the author pointes out the assumed conditions that translate the system of man-machine-environment into the system of man-technology-organization, and illustrate the course that establishes MTO model, in the meanwhile, the human error classification model of MTO is established based on the socio-technical approach, Thus we can probe into the error of organization and technology error which hides the back of individual human error. Through finding, organization error has 5 big categories, that is supervision error, information error, training error, surrounding environment error and work condition error; The technological error has 3 big classes, e.g. the error of tool and equipment, procedure error and the error of man-machine interface. Every big class has further subdivided into several concrete errors so that the lessons are drawn from these errors. At the same time, the possibility and necessity of collecting organization & technology error data is analyzed in detail, and their data collection method of is also discussed in the paper. 8 standardized forms of data collection are listed, and some measures is pointed out which should be taken in the course of data collection so as to make the work of data collection standardize and offer the condition for the...