Forecasting construction time for road projects and infrastructure using the fuzzy PERT method

Abstract

It is essential to organize and plan any project at all stages, from start to finish; however, it is also necessary to certify completion by means of correct and effective methods. Scheduling projects using scientific approaches and sophisticated scheduling tools helps guarantee to project completion by the scheduled dates, as well as allowing institutions and municipal departments to make accurate estimations for project execution. The construction of roads includes several activity packages, one of which is the generation of road infrastructure such as that related to the sewer work (such as water and wastewater), the drilling and pouring work, the electrical work, and finally the paving. One of the most difficult jobs facing a manager in terms of such a project is the scheduling of activities over time. In many projects, the scheduling is done compactly, such that each activity that matches one scheduled in previous projects is scheduled but not detailed, with estimates based on previous overall times. In this research, however, each major activity was divided into several sub-activities to allow more accurate calculation of duration based on several criteria that affect the activities’ use of time. Three timings were thus estimated for each activity: a pessimistic time, an optimistic time, and a normal time. The PERT method was improved by using a Fuzzy Delphi technique to estimate the actual time required to implement each activity. The overall scheduling of a project in Karbala city, the rehabilitation of the road connecting the intersection of the Al-Ghadeer district to the intersection of the Al-Mudraa district was done in this manner. The project implementation period was predicted at 301 days, which is very close to the 322-day actual duration in which the work was carried out. The findings confirmed that the improved PERT method using the Fuzzy Delphi technique was the most accurate way to predict execution time.