| Summary: | Current approaches to evaluating the success of construction projects are inadequate in assessing and improving the management of ongoing and future projects. This is concluded from a study of the relevant literature and of construction project management in Sri Lanka and the U.K. The main criticisms of current practice are the ad-hoc nature of the evaluation itself, the failure to properly relate the evaluation: to the original project objectives; to contextual conditions; and to realistic targets derived from historical databases of similar projects. The absence of any such comprehensive database is a particular shortcoming in Sri Lanka. This research work develops and tests a framework for a comprehensive construction project evaluation system that would overcome such shortcomings.The system would help an evaluator generate a specific project evaluation framework depending on the category (type) and stage of the project and the particular purpose of the evaluation. Such a framework would contain criteria of project performance (eg: related to cost, time, quality, safety, satisfaction etc. and associated sub-criteria), hierarchies of 'indicators' by which to measure such criteria and typical target values (and ranges) of such indicators in that category. Such target values are weighted by the specific project profile, based on project priorities and contextual conditions. The sub-systems of the proposed evaluation system are the criteria, the indicators (with typical target values) and an expert system front-end. Supporting 'tools' modules contain techniques for information elicitation, weighting, analysis and presentation. The three sub-systems of criteria, indicators and the expert system were tested. For example, groups of cost indicators and their average values within specific project categories were derived from data obtained from 138 building, 35 roadworks and 38 bridge projects in Sri Lanka. Their validity was tested by their consistency within project categories, by comparison with projects outside the original database and at a Workshop with experts in Sri Lanka. A pilot expert system was developed to demonstrate the viability of integrating appropriate modules. It is concluded that the proposed system is viable and would provide comprehensive and realistic evaluations of the management of construction projects.
|
|---|
