Genetic Algorithm for Resource Levelling Problem in Construction Projects

Abstract

The resource levelling problem (RLP) arises when with sufficient available resources, it is mandatory to have minimum fluctuations in the pattern of resource usage. These fluctuations are very much undesirable because it causes problem in the utilization of labour and monetary concern for the contractor. This kind of problem aims at minimizing total duration or making span of a project-dependent to precedence relations between activities and finite resource availability. A resource levelling approach gives direction to the person doing scheduling to plan activities in a project with smooth resource utilization without exceeding the constraints of resource availability. It is also often useful in minimizing the project duration within limited availability of resources for making project profitable. In RLP, start times of noncritical activities are shifted within their available float times so that the resource usage profiles are levelled best. Existing research has done to analyse various methods and algorithms for the resource levelling problem. In this study, we have discussed the scheduling process of RLP using genetic algorithm technique which is evaluated by a sample problem where resource levelling index (RLI) is improved by 48% to conventional schedule resource profile. Analysis of this study shows that GA could be an effective asset for getting optimized solutions in construction industry. KeywordsResource levellingGenetic algorithmResource managementProject scheduling