Schedule Risk Assessments Using a Precedence Network: An Object-Oriented Bayesian Approach

Abstract

Uncertainties affect the scheduling of construction projects. They are among the main causes of delays in construction projects, which increase project costs and reduce project quality. Uncertainties in scheduling affect the duration of activities. Under such conditions, the duration of each activity is influenced by several risk factors; however, previous Bayesian scheduling models considered the effects of one or no factor on the duration of tasks. Also, all parameters of the scheduling network were not considered in those models. In this study, given the causal relationships between uncertainties and their multiple effects on the task duration, a duration model is presented. Next, the effects of uncertainties on all scheduling parameters are applied to Bayesian object-oriented networks through mapping precedence networks. Object-oriented Bayesian networks are a new project management approach to risk assessment and uncertainties in decision-making. This approach is quite effective in considering the causal structure of risk variables and assessing the repetitive structure of precedence networks in large-scale scheduling networks. The project completion time, total slack, delays, as well as other scheduling parameters are estimated using the model proposed, which is titled the Bayesian Precedence Network.