DESIGN OF CLOUD-BASED INTELLIGENT PLATFORM FOR INCIDENTS MANAGEMENT & PREDICTION USING MACHINE LEARNING

Abstract

Advances in data collection and storage infrastructure present a once-in-a-generation chance to integrate both data and emergency resources in each location into a dynamic learning system capable of anticipating and swiftly responding to situations involving different types of hazards. Currently, incident details are not always thoroughly documented and may be described in a variety of formats. Consequently, finding major issues in accident reports is a time-consuming and error-prone process. We tend to respond to the incidents manually and in a different way for the same exact incidents, even if the incidents have happened before. It makes the analysis task more difficult, affects the correction deadline of incidents as well as involves a high risk in terms of financial and workers' well-being. This paper discusses the design of an online incident report by exploiting the advances in cloud computing and the Internet of Things (IoT). This system will support incident analysis and incident prediction using collected incident data. Various strategies in the literature focus on assessing attributes and forecasting occurrences for specific conditions, but when used in a more significant and generic scenario, these models are ineffective. To account for these characteristics, we propose the use of machine learning (ML), specifically Similarity-Based Agglomerative Clustering (SBAC) analysis, to cater to various types of incidents of a generic nature. The proposed platform's principal objective is to predict the probability of the occurrence of a hazardous incident and execute rectification actions before the incident occurs. It can help organizations predict the likelihood of a fatal incident or accident occurring in the near future by analyzing historical data. This solution will help to improve overall safety ratings, as well as employee health, cost savings, and new project growth.