A geometric interpretation of diagnostic data from a digital machine: Based on a study of the morris, illinois electronic central office

Using the diagnostic data collected for the Morris Central Control malfunction dictionary, we devise a natural concept of “distances” between malfunctions. Ten thousand malfunctions were placed as points in six-dimensional space in such a way that the Euclidean interpoint distances approximately equaled the diagnostic “distances“. The remarkable fact that this is possible has many implications. By finding circuit characteristics common to a cluster of neighboring malfunctions, we are able to associate these characteristics with the region of the six-dimensional space which holds these malfunctions. By this means, we characterize various regions of space according to functional troubles. This suggests a technique for locating malfunctions and also suggests some longer-range possibilities.