|Abstract or Summary
- This dissertation is an attempt at developing a method for the analysis and estimation of the effects of disruptions due to uncertainties. Such uncertainties may result from design changes and engineering drawing delays in large-scale, complex, research and development, or construction projects.
In order to provide management with a simple management tool for the analysis of disruption problems, the following works have been done in this dissertation. I
nformation Theory is used to provide an entropy formula and the capacity concept for a project, based on a transformation function. A Systematic prediction procedure is developed. The procedure can provide management with warning information on the likely trouble spots in the early stage of a project as well as during project execution.
A hypothesis is advanced that most of engineering drawing delays follow an exponential distribution. Using the maximum entropy principle, an exponential distribution is shown to be an adequate probability distribution for random occurring drawing delays. Furthermore, to support the hypothesis, x² 'goodness-of-fit' tests are conducted for actual industrial data concerning drawing delays.
An entropy formula of a triangular distribution for activity is derived. The entropy formula is used as a working vehicle for developing an information processing capacity estimation method for a project. In addition, an entropy conversion method is proposed to compare and choose project(s) among more than one project that has different measurement units for activity duration, in terms of uncertainty of the project completion date.
An information processing capacity estimation method is developed. The method can estimate the capacity of a project to handle equivocation due to design changes. It can also identify the associated activities and the necessary amount of human resources, considering the project completion date with 50% chance of success. Furthermore, the method can estimate the project slippages when the capacity is not large enough to handle the equivocation.
In an attempt to evaluate the predictive ability of the method, project slippages estimated by the information processing capacity estimation method are compared with the results obtained by a computer Monte Carlo simulation program, CRASH. A x² 'goodness-of-fit' test is conducted. The result of the test shows that the estimated project slippages do not significantly differ from those obtained from the CRASH computer program at the 0.005 significance level. We conclude that the information processing capacity estimation method may be suggested as an expedient means of evaluating project status for management in the different stages of project execution. The effect of fatigue on the capacity of a project is evaluated. In one example, it is shown that a 25% increase of total working hours over the base schedule (6-day week, 8-hour day) results in only 16% increase of the capacity of the project. The effect of fatigue is shown as a 35% time lost of the total increased working hours.