Artificial intelligence for predicting heat stress in climate sensitive environments based on a systematic literature review

The escalating issue of thermal stress, exacerbated by global warming, poses significant long-term challenges to human health and ecosystems. This study investigates the role of artificial intelligence (AI) in addressing heat stress induced by climate change, with a focus on advanced predictive capabilities and mitigation strategies. A systematic literature review was conducted using prominent databases, including PubMed, Scopus, Web of Science, and Google Scholar, to identify relevant studies on the intersection of climate change-induced thermal stress and AI applications. Out of 608 studies reviewed, 14 were selected for in-depth analysis. This study offers a comprehensive overview of AI and machine learning (ML) methodologies employed to monitor and predict heat stress events associated with climate change. Notably, it highlights the growing focus on hybrid AI models, which integrate multiple approaches to improve the accuracy of heat stress predictions. Key areas of prediction include body and skin temperature, heat stroke, and environmental indices such as WBGT, MWD, and TSENS. The review also underscores the potential of AI and ML algorithms in forecasting both heat stress risks and broader global warming trends. Hybrid AI models tend to demonstrate higher predictive accuracy compared to standalone approaches within the reviewed studies. Across the reviewed studies, hybrid and ensemble AI models generally demonstrated improved predictive performance compared with standalone approaches, particularly when integrating environmental and physiological variables. However, due to heterogeneity in study design, datasets, and evaluation methods, direct comparison of performance metrics across studies was not feasible. Overall, the findings highlight the potential of AI-based approaches to support early warning systems and occupational heat stress management in the context of climate change.