Novel hybrid model for addressing uncertainty of the road safety composite indicator: integration of DEA and weighted GRA

Creating a composite indicator is a popular concept in evaluating and comparing road safety of territories. While many other approaches are used, most current interest is on Data Envelopment Analysis (DEA) for measuring the relative performance in efficiency terms. Usefulness of DEA, which measure performance in efficiency terms, has already been proven. However, the indicators commonly used to construct a road safety composite index are not always precise and accurate (particularly safety performance indicators), and the results obtained from this type of data does not seem to be solid. So, the main aim of this paper is to represent the novel methodology to effectively evaluate the state of road safety and create a reliable composite index of the selected entities when imprecise data are involved and, to produce reliable composite index under uncertain environment. DEA, the weighting method Fan–Ma, and the Grey Relational Analysis (GRA) were integrated into hybrid methodology to obtain a more realistic and relevant picture of road safety. Applying this hybrid methodology, peculiarity of DEA is retained, scores are further differentiated, and entities are ranked and classified according to the road safety level. A case study was conducted to evaluate and rank municipalities, determine road safety classes and benchmark the territories under study. Results were verified indicating the robustness and effectiveness of the proposed methodology and its superiority to basic DEA as regards the territory ranking. First published online 3 February 2026