MODELAGEM DE CRONOGRAMA DE PROJETO PELA FERRAMENTA DSM COM APOIO AO GERENCIAMENTO E TOMADA DE DECISÕES PELA SIMULAÇÃO DE MONTE CARLO

Abstract

Delays in construction projects are a recurrent fact in several countries and regions for the most varied types of works. These delays, in addition to negatively impacting the image of companies contracted in the provision of construction services, cause several financial losses to interested parties: contractor and hired. Several researches that address the factors that lead to delays in the work schedule indicate the variation in the duration of the activities and, or the lack of an inadequate estimate for the duration of the activities, among the ten major delay factors in the projects. They contribute to the inadequate estimation of production of the activities, the lack of knowledge of the complexity of the activities, and also the lack or insufficiency of historical data that allow a safe estimate of the duration of the activities. Thus, this study aims to describe the behavior of variability in the duration of activities in an environment with little activity productivity data. The productivity data analyzed were taken from three similar construction projects carried out by the same company and which were provided as case studies of this work. For the analysis of the variability and description of the behavior of the activities, after the data collection of productivity of the project activities, the non-parametric Bootstrap data resampling associated with Monte Carlo Simulation (MCS) was used as techniques. Later, in order to verify the reliability of the results of variation in the duration of the activities, the schedule of the case studies was modeled using the Critical Path Method (CPM) and the Dependency Structure Matrix (DSM); and submitted to MCS. The simulated results of the total duration of the projects were adequate for the actual completion period of the projects studied, leading to the conclusion that the results of variation in the duration of the activities obtained by the cited technique are reliable.