Comprehensive structured analysis of machine learning in safety models

Abstract

Machine learning (ML) integration into various industries has revolutionized operations recently, enhancing efficiency and predictive capabilities. However, the rapid adoption of ML models also presents significant safety concerns that are highly demanded. To achieve this, scholarly articles from reputable databases such as Scopus and Web of Science (WoS) focus on studies published between 2022 and 2024, which were extensively searched. The study's flow is based on the PRISMA framework. The database found (n=40) that the final primary data was analyzed. The findings were divided into three themes: i) safety and risk management, ii) ML and artificial intelligence (AI) applications in safety, and iii) smart technology for safety. The conclusion highlights the need for continuous monitoring and updating of the safety protocols to keep in step with the growing ML landscape. This review contributes to the understanding of ML safety. It offers global lessons that can guide future research and policy-making efforts to ensure ML technologies' safe and ethical use.