This research refines and validates the Network-Wide Road Safety Assessment (NWRSA) methodology for Italian regional roads by closing key methodological gaps and demonstrating reproducible, scalable approaches to network-level safety appraisal, through its implementation in an information system. The work combines a systematic review of automated and manual road-assessment methods with empirical analyses that convert NWRSA judgment calls into measurable protocols and operational thresholds. Pilot studies on representative Italian regional roads will validate measurement accuracy, compare fully automated and human-in-loop implementation strategies, and produce a reference population for reactive assessments. Expected contributions include operationalized NWRSA rules, validated modelling of attribute–crash relationships, and practical guidance for regional agencies on implementing reliable, defensible network assessments.
*********
Context and Problem: The European TEN-T road network is among the safest globally; however, the majority of crashes occur on lower order roads. Current EU Directives on Road Infrastructure Safety Management (RISM: Directive 2008/96/EC and Directive 2019/1936) establish regulations to enhance road safety. Yet, the official implementation guidelines for Network-wide Road Safety Assessments (NWA) focus on motorways and primary roads, leaving secondary roads—where a disproportionate share of crashes occur—insufficiently addressed.
Research Aim: My PhD research aims to make the implementation of the current NWA methodology easier to implement by developing an automated decision support for road safety assessments and cover several implementation gaps in the methodology. In addition, the adaptation and extension of the methodology to cover secondary and/or regional roads will be carried out.
Current Activities and Progress: To date, my work has centred on an extensive literature review of network-ranking methodologies and their implementation. The specific focus has been to identify and prioritize the road infrastructure features—such as geometry, —that most significantly influence accident rates, and their prediction, on secondary roads. The findings from this review will be used to define the parameters for the new assessment model.
Publications as indexed by Google Scholar
In progress ...