PSI - Issue 84

Galileo Tamasi et al. / Procedia Structural Integrity 84 (2026) 725–732

729

hail), road surface conditions (skid resistance, roughness, and bearing capacity), and the state of traffic control devices such as horizontal markings, vertical signs, and luminous signals. Such a data-driven approach is consistent with the latest trends in risk-based asset management, which prioritize interventions based on empirical probability and environmental risk factors (Dehghani et al., 2025; Ouyang, 2021). To operationalize this vision, the research adopts a quantitative risk assessment framework inspired by the rigorous safety standards of the aviation industry (Transportation Research Board - National Academies of Sciences Engineering and Medicine, 2011a). This framework is structured into three interdependent modules: a probability of occurrence analysis, a spatial localization model, and a consequence and harm model (see Fig.2).

Fig. 2. Three-Part assessment framework for airfield risk. Source: (Transportation Research Board - National Academies of Sciences Engineering and Medicine, 2014).

The probability module utilizes logistic regression to quantify the likelihood of incident precursors based on the aforementioned infrastructure and meteorological variables (Transportation Research Board - National Academies of Sciences Engineering and Medicine, 2011b). The spatial localization module then applies two-dimensional Gaussian models to determine the probable distance from the road axis following a loss of control, utilizing extensive research on veer-off and overrun distributions (Transportation Research Board - National Academies of Sciences Engineering and Medicine, 2014, 2008). Finally, the harm model evaluates the potential for loss of life and property damage by correlating the localized probability of impact with the specific obstacles present on the road segment and the vulnerability of users (Transportation Research Board - National Academies of Sciences Engineering and Medicine, 2016). This analytical process is integrated into a broader decision-making framework that emphasizes evaluating multiple intervention options to ensure that safety countermeasures effectively reduce systemic vulnerability (National Research Council, 2009). 3.3. Enhancing reliability through predictive analytics Integrating these advanced sensors and communication protocols enhances the overall reliability of the road system (Bazhanov, 2019). Rather than waiting for a crash to occur to justify safety improvements, the proposed model utilizes the frequency and severity of separation reductions to evaluate the performance of a work zone mobility plan in real time. This methodology aligns with the data needs of modern highway agencies, moving away from "business as usual" toward an evidence-based safety culture (Nlenanya and Smadi, 2018). By correlating vehicle behavior with infrastructure condition data - such as sign support integrity or pavement quality - the operator can implement targeted