Mitigating Psychosocial Risks in Construction Through Predictive Machine Learning Analytics

Abstract

The construction industry is one of the most hazardous sectors globally, characterized by high rates of workplace accidents and injuries. While physical risks are extensively addressed, psychosocial risks—such as stress, job insecurity, and poor organizational culture—remain underexplored yet critical contributors to unsafe practices and accidents.

This paper investigates the correlation between accident occurrence on construction sites and psychosocial risks, emphasizing how factors like workplace stress and organizational deficiencies lead to unsafe behaviors and heightened accident rates. Following this analysis, the study explores the role of predictive analytics and Machine Learning (ML) models in identifying and mitigating these risks to enhance workplace safety.

By analyzing historical data, survey responses, and environmental factors, this study highlights how advanced predictive tools can uncover patterns and correlations that traditional methods often miss. The findings emphasize the potential of ML-driven approaches to proactively address psychosocial risks, reduce accidents, and promote a culture of safety. This research, therefore, underscores the need for integrating technological innovations with human-centric risk management strategies to achieve a safer construction environment.