DEVELOPMENT OF AN OPTIMIZED SCHEDULE SYSTEM FOR REPETITIVE CONSTRUCTION PROJECT MANAGEMENT USING GENETIC ALGORITHM

Abstract

The high rate of construction project abandonment in Nigeria had led to researches into designing various systems of organizing the construction process for efficient project delivery. This research looked into how Industrial Technology (I.T) generated solutions could be augmented into solving the recurring delay in project delivery with its high indirect cost resulting from interruptions in workflow of resources in a repetitive construction project site. In this work, an optimized schedule system for repetitive construction project using Genetic Algorithm (GA) was developed. The constraints used in the Optimized Schedule System design was selected through a questionnaire survey among one hundred and forty-one (141) participants across three states in Nigeria. This brought about sixty-three (63) cases/scenarios that occur on an estate construction project site which were placed into seven (7) constraint groups. The effects of these scenarios were rated separately to evaluate their severity to causing late project completion and cost overrun through a 5-point Linkert scale ranging from Strongly influence to Virtually no influence. A total of one hundred and thirty (130) valid responses were collected and analysed. The result showed the constraints as Productivity constraint, Precedence constraint, Resource constraint, Quantity of work constraint, Options constraint, Price fluctuation constraint and Project environment constraint with a Friedman’s mean significance rank of 5.31, 5.22, 5.10, 3.74, 3.27,2.87 and 2.50 respectively for project delay. For cost overrun, it gave Productivity constraint, Resource constraint, Quantity of work constraint, Precedence constraint, Price fluctuation constraint, Options constraint and Project environment constraint, with a Friedman’s mean significance rank of 4.89, 4.85, 4.62, 4.40, 3.48, 3.21 and 2.54 respectively. A Schedule System Design was then developed using binary coded GA programming in an excel sheet containing data from Omoku estate construction site, arranged to show the constraints discussed in the questionnaire in real site form. The result obtained from the first optimization involving minimizing total project cost which was ₦408,868,140 at 156 days, when compared with the actual(validated) project completion time and cost of ₦450,867,620 and 150 days respectively, produced an extension in project duration by 6 days and savings of approximately fourty-two (42) million naira. However, a second optimization involving minimization of both total project cost & time produced a cost of ₦364,996,420 and duration of 136days which when compared to the contractual duration and cost of ₦456,843,568.00 and 160 days, produced a reduction in project duration by 25 days and savings of approximately eighty-six (86) million naira. This research established the fact that introduction of Engineering system design such as Genetic Algorithm into the scheduling of construction project would aid the managers in managing resources towards efficient project delivery at low cost