Mere collection of failure information and accident records do not effectively relay the knowledge associated with the case to the reader. We propose to collect data in a structured manner so the message is better transferred to the information receiver. We further developed a scheme that records the essence of each failure case in a sequence of predefined phrases displayed to the recorder in a hierarchy of phrases. We call the sequence the “scenario” of the event. Arranging the phrases in descending steps and supplementing it with an illustration and a key knowledge sentence composes the visual summary of the case. A glance at the visual summary and reading the scenario steps generate a good image of the case in the receiver’s mind. Among the predefined hierarchical phrases, we call those that express the cause of the event, failure cause phrases. Recording high level failure cause phrases from the hierarchy forces the event recorder to evaluate the root cause of the failure. To the top and second level (in the phrase hierarchy) failure cause phrases, we assigned 5-space vector components to characterize each phrase in terms of “knowledge”, “carefulness”, “judgment”, “organization”, and “nature”. This vector characterization of the failure cause phrases with the scenario allows us to further characterize each failure case as a linear combination of the predefined phrases. Once each failure case has its vector characterization, we can evaluate its similarity with other cases. Also, if we find the characterization of an individual, group, or organization in the same 5-space, we can warn about failure cases with similar characteristics that are likely to happen. The method is powerful in predicting failures so they can be avoided before happening.

This content is only available via PDF.
You do not currently have access to this content.