Analyzing and Applying Nonparametric Algorithm to Evaluate Safety Equipment

Abstract

This research proposes a new methodology in handling of safety equipment data from construction companies to determine factors influencing equipment. Project safety management is an important subject and one that interests researchers, practitioners and decision-makers.  Yet few of them employ the cluster detection algorithm method tools created and tested in other disciplines.  The core question is a matter of the quality of conclusions that one can draw from available data and measurements by employing new cluster detection algorithm methods. A secondary problem is the complexity in applying these methods, as well as how applicable the results obtained are. Real building data for six years is obtained for the experiments using the safety equipment dataset of a leading Libyan company. The new algorithm, NCA, is superior to the current outlier mining strategies. He helped in making necessary decisions to ensure the company's safety equipment. The NCA mining algorithm can identify the malfunctioning safety equipment and rank them according on how abnormal they are. This helps the equipment manager find hidden issues related to the maintenance of safety equipment. By comparing irregular safety equipment with other safety equipment in the same group of equipment, a manager may thus make the best decision when managing the equipment, including buying, replacing, repairing, or excluding it. The project's applicability of these approaches as opposed to established methodologies, such as the Formal Safety Assessment (FSA), is addressed in the conclusion of the paper.