Abstract:
Public infrastructure projects selection is critically important for the government to
successfully achieve its strategic directions and improve its contribution to the national
economic and social development. Due to the limitation of resources and the vigorous
changes in internal and external environment, infrastructure projects selection is a
challenging task for decision makers. The Ministry of Works is the construction arm of
Kingdom of Bahrain government and it is entrusted with infrastructure projects like roads
constructions and sanitary engineering projects. Despite the availability of various tools for
project selection, unless the Ministry of Works decides on its own tailored decision criteria
for project selection, then all the project selection tools will add no or little value.
Generally, these theoretical frameworks prove to be cumbersome for governmental
organizations, in which decision makers are looking for a practical yet effective tool that
can be incorporated seamlessly in their operations. Therefore, the thesis intended to
investigate the decision criteria with the most significant impact on the Ministry of Works
infrastructure project selection. Then, to develop and propose an effective method that can
be beneficially adopted by the Ministry of Works. A questionnaire and semi- structured
interviews were designed to investigate what decision criteria are currently practised and
which criteria is considered of significant importance. The study identified the decision
criteria factors that are considered important for the Ministry of Works project selection.
The decision criteria were then fed in a Fuzzy Inference System (FIS) using MATLAB to
construct an expert system for infrastructure projects selection. Two proposed FIS models
were developed, both for project selection and prioritization. The first model is designed
with one output to give an overall score for projects importance. The second model is a
graphical two-output model which is intuitively represented on a graph. It scores the
projects based on two overarching criteria and then the projects are plotted on the graph
for selection and prioritization decision. The study concluded that given the right criteria
the proposed models can be utilized to select and prioritize projects. The recommendations
support the necessity to implement scientific based practices like fuzzy logic in public
infrastructure project selection.