Non-radial DEA model: A new approach to evaluation of safety at railway level crossings

Railway level crossings (RLCs) are critical points characterized by a large number of accidents per year due to the intersection of railway and roadway infrastructures. The improvement of safety at RLCs is of crucial importance worldwide. This paper proposes a new approach for evaluating safety at RLCs using non-radial DEA (Data Envelopment Analysis) model. The introduced non-radial DEA model is employed for the evaluation of railway efficiency of European countries regarding the level of safety at RLCs through considering desirable and undesirable inputs, as well as desirable and undesirable outputs for the period from 2010 to 2012 and for the year 2014. Due to missing the comprehensive set of data for roadway traffic, the evaluation of safety at railway RLCs has been conducted from railway viewpoint. Through the application of modified non-radial DEA model, results indicate the most efficient countries, as well as countries with a lower efficiency of railway performance in terms of the level of safety at RLCs. In order to check the validity of modified non-radial DEA model, comparison of results and sensitivity analysis were conducted. Despite the validity, results of the sensitivity analysis indicated also some weaknesses of the modified non-radial DEA model related to missing and inaccurate data, the number of variables included, and the selection of inputs and outputs. A modified non-radial DEA model, as proposed in the paper, with overall set of data can be used for evaluating the efficiency of safety improvement at RLCs, efficiency of different countermeasures before and after implementation at RLCs of different.