A Decision Model for Human Error in the Design and Construction Process.

Modern structural reliability analysis is concerned with the uncertainty that affects structural safety and serviceability. Current reliability methods model uncertainties caused by natural variations of loads and resistance. These methods assume that these variations are within certain tolerances or acceptable limits. However, surveys have shown that the major cause of structural failure, in both ultimate and serviceability limit states, is not the natural variation of loads and resistance but the out-of-tolerance variations caused by human error during the design and construction. A decision model that incorporates the effects of human error into structural reliability analysis was developed based on the probability of errors remaining in the final structure after completing the entire design-construction process. It also considers multiple failure modes, multiple design parameters, and cost of failure. The activities in the design-construction process were characterized into correction facet and action facet. The model presented here will serve as the basis for selection among design-construction process alternatives. The selection criterion is based on minimizing expected cost of failure through the consideration of the probability of making errors, sensitivity analysis for errors in design parameters, probability of failure, the estimated cost of the alternative selected, the expected cost of correction, and the expected cost of failure. and , it is expected to assist engineers in selecting among design and construction alternatives to increase reliability at reasonable cost. The model is demonstrated for a simple structure to show the method of selecting among several process alternatives at different levels and stages of design and construction.PhDCivil engineeringUniversity of Michiganhttp://deepblue.lib.umich.edu/bitstream/2027.42/161434/1/8712174.pd