An expert system for strategic control of accidents and insurers’ risks in building construction projects

Access & Terms of Use
metadata only access
Abstract
Building construction projects appear to have higher accident rates. Contractors procure workers’ compensation insurance (WCI) to transfer these risks to insurance companies. The commitment of insurers under WCI is extremely broad; there are no exclusions and ceilings on their liabilities. They must quote adequate premiums to cover future risks. Yet, the prices should be low enough to penetrate the market. Thus, accomplishing rigorous risk and market assessments to decide optimal premiums for building projects is crucial. Traditionally, experience rating approach has been adopted for WCI premium-rating. However, this approach has been found ineffective for construction. Hence, the purpose of this study is to develop an effective WCI premium-rating model for building projects, and to automate the model as an expert system. A new WCI premium-rating model was developed based on the findings of a literature review and a questionnaire survey. A hybrid of interviews and past workers’ compensation claims data analysis was pursued to develop the conceptual model of a fuzzy expert system to automate the proposed model. It was then prototyped and verified with Turing tests. The proposed expert system advocates real-time assessments of project hazards, safety, market conditions and insurers’ internal factors for premium-rating. It also establishes an effective risk control strategy via a well-structured incentive system for contractors and clients. Its implementation in the insurance industry can curtail accidents in the construction industry, thereby minimising insurers’ financial risks.
Persistent link to this record
DOI
Additional Link
Author(s)
Imriyas, Kamardeen
Supervisor(s)
Creator(s)
Editor(s)
Translator(s)
Curator(s)
Designer(s)
Arranger(s)
Composer(s)
Recordist(s)
Conference Proceedings Editor(s)
Other Contributor(s)
Corporate/Industry Contributor(s)
Publication Year
2009
Resource Type
Journal Article
Degree Type