AI-Driven Predictive Safety Analytics: Enhancing Workplace Security

1. Introduction

According to Rasmussen (1998), the fundamental cause of accidents often lies in human errors committed by individuals directly involved in the unfolding events, with statistics indicating that 70-80% of industrial accidents stem from such errors. Human errors are correctable through behavioural interventions, underscoring the vital role of behaviour in safety. Guldenmund (2000) emphasizes that safety processes should consider three key domains: environment (including equipment and management systems), person (encompassing employee knowledge, skills, and motivations), and behaviour (involving compliance, recognition, communication, and active care). Behaviour emerges as the primary tool for survival, especially when other safeguards fail (Galloway, 2012). Galloway (2012) further contends that in the absence of proper tools or systems, workers rely on their behaviour for self-preservation. Therefore, enhancing workers' safety behaviours offers a promising avenue for reducing human errors and elevating safety at the organizational level. Parboteeah and Kapp (2008) highlight an overlooked connection between workplace safety and ethics, with only two studies delving into this association. In the first study by McKendall et al. (2002), they investigated how components of an ethics program, including ethical codes, communication, training, and integration into human resources practices, related to Occupational Safety and Health Act (OSH Act) violations. Surprisingly, the findings suggested that ethical compliance programs might be used to divert attention from illegal activities rather than fostering legitimate conduct. The second study by Parboteeah and Kapp (2008) introduced the novel concept that an organizational ethical climate plays a pivotal role in enhancing workplace safety, challenging the conventional contingent reward approach that relies on incentives and penalties to promote safety behaviours. This study aims to comprehensively explore how AI-Driven Predictive Safety methods are utilized within organizations and what factors influence employee engagement with these methods. It seeks to conduct a comparative analysis of these aspects across private and public sector organizations through a systematic review of existing literature. Despite a growing body of research on methods and factors enhancing AI-Driven Predictive Safety in various sectors and generations, there is a notable absence of a comprehensive synthesis and conceptualization of these findings. Therefore, this research addresses the fundamental question: What methods and factors are prevalent in the Information Systems (IS) literature for enhancing employees' AI-Driven Predictive Safety across both private and public sectors?