This paper aims to demonstrate nonlinear risk factor interactions based on a data-driven approach using a Bayesian network model, providing a road safety use case. Road safety is a critical issue worldwide, with approximately 1.3 million road traffic deaths each year (WHO). Traditional road safety risk assessment methods often analyze individual factors separately; however, these assessments fail to capture the complex dynamics of real-world analysis, in which multiple factors interact through nonlinear relationships. In this study, a novel road safety risk assessment approach that uses a Bayesian network model to explore the nonlinear relationships among road safety risk factors is developed. Through the analysis of extensive crash reports from the state of Maryland, the complex interdependencies among various risk factors and their cumulative impact on road safety are investigated. Our findings show that two combined risk factors have different effects on risk level when considered individually. Two case studies related to human state risk factors and environmental risk factors, such as driving under the influence and snowy roads, as well as fatigue and snowy roads, have an amplified effect on the risk level. The findings highlight the importance of considering nonlinear interactions among risk factors when developing effective and targeted strategies for accident prevention and road safety improvement. This research contributes to the field of road safety by presenting a new methodology for understanding and mitigating road safety hazards.

Road safety is a priority, worldwide. The European Commission aims to reduce fatalities by 2030. The same goal was set for the US. These goals stem from the World Health Organization's (WHO's) broader global context, which has distinctly emphasized a substantial reduction in road traffic injuries. Although different risk factors were observed in different geographical locations, the major risk factors for all locations were similar. They involve influencing human behavior, such as speeding or driving. Several methods have been used to better understand and extract risk factors. However, the complexity of road traffic implies the need for a multi-criteria method. As a result, the analytical hierarchy process (AHP) has emerged as a potential method for this type of risk. The AHP is commonly associated with the use of qualitative methods such as surveys. We propose a novel semi-quantitative multi-criteria risk model (SMCRisk) based on the AHP, deployed in a quantitative and partially qualitative manner by adding a severity factor. The multi-level framework differentiates between the driver's behavior and the driver's state. Our method results correspond to a real situation and confirm that driver behavior and state are major risk factors. In future, this method will lay the foundation for integrating a fully quantitative method by considering the potential use of data originating directly from the IoT, which is a part of our research on holistic risk assessment.

In complex systems, understanding the nonlinear interactions among risk factors is essential for accurate risk analysis. However, traditional linear models often fail to capture these complex interdependencies, leading to significant gaps in risk prediction. The aim of this study is to present a novel approach for risk analysis of nonlinear risk interactions using Bayesian networks (BNs), thereby providing a broadly applicable method for risk management and mitigation. Specifically, this study applies a BN-based framework that integrates conditional dependencies and nonlinear effects to illustrate how multifactor risk interactions operate synergistically. Using a step-by-step approach, the interactions among multiple risk factors are first mathematically formalized, and then this framework is applied to a case study of road safety using crash report data. Additionally, a second validation case in public health (type 2 diabetes risk) is included in supplementary materials to illustrate the broader applicability of the framework. The findings demonstrate through BNs and a mathematical framework, how to analyse complex interactions more accurately than traditional methods can, revealing the amplifying or mitigating effects of individual risk factors on outcomes. This approach offers more accurate risk representations and is applicable not only to road safety but also to complex environments, such as healthcare and environmental risk analysis.