Productivity analysis in construction using an expert system approach

The purpose of the work described in this thesis is to computerise a productivity causal research model using knowledge based expert system techniques. To achieve this a microcomputer based prototype total system has been developed. The system enables construction managers or engineers to analyse and diagnose construction site problems related to productivity that occur in the course of the excavation process in order to establish their causes. The system elicits information relevant to the problems by asking questions. It then examines the problems presented by construction managers or engineers, applies its set of inference rules, determines probable causes of the problems, and then produces advice on them. The knowledge structure of the prototype total system is in the form of causal links that interconnect variables which represent managerial factors affecting system productivity in construction. The system employs the production rule-based representation techniques to represent the knowledge base. The control structure of the system uses the effect and cause mechanism to determine a solution. The system consists of 11 independent programs. The knowledge base contains 58 facts and 319 rules. The results of this system have shown that the system puts theoretical knowledge into practical use and provides a useful tool to educate and train inexperienced engineers or students in the area of productivity analysis in construction.