One technique used to improve highway safety from the point of view of the infrastructure is to examine the consistency of the design. Design consistency refers to if highway geometry is conformance to driver expectancy. When the consistency of the road is inadequate, the more likely it is that drivers will be startled and a crash will occur. The consistency, based on operating speed, has been calculated in Spanish two-lane rural highways. This consistency has been evaluated using a local method, to measure the consistency of each element of the road and using a global method, to measure the consistency of a segment of the road. Different models of consistency have been compared using Decision Trees (DTs). DTs are a Data Mining Techniques which can be used to solve classification problems. The results show that DTs are a suitable technique to compare consistence models and they permit to establish limits between the different models analyzed.

En los últimos años, la Universitat Politècnica de València (UPV) ha promovido, dentro de su programa de innovación educativa “Aprendizaje y Docencia” (A+D), la puesta en marcha de Proyectos de Innovación y Mejora Educativa (PIME). Durante el curso 2019/2020 se ha implantado un PIME basado en Aprendizaje Basado en Proyectos (ABP) que reúne a dos asignaturas consecutivas de tercer curso de Ingeniería Civil. En este artículo se muestran los resultados alcanzados en el primer semestre de implantación. Los alumnos han expresado claramente su interés y preferencia por este tipo de metodología. Han demostrado haber adquirido un aprendizaje en profundidad y ser capaces de adquirir conocimientos de forma autónoma, así como transferir estos a la realidad de la profesión mediante el desarrollo de su capacidad de trabajar en equipo y la posibilidad de realizar visitas de campo donde los conocimientos recibidos adquieren un carácter práctico. Además, han desarrollado las competencias esperadas, integrando aspectos académicos con otros sociales y éticos.

In order to facilitate the management of the results obtained in the project "Analysis of the relationship between Road Safety Deficiencies, crashes and hazardous sections" financed by Public Works Agency of the Regional Government of Andalusia (AOPJA) and led by the research group TRYSE from University of Granada, a safety management tool has been developed. This application allows safety managers to consult some factors affecting crashes on two-lane rural highways. The main aim of that project was to analyze the influence of some road deficiencies on crashes and hazardous sections in the Complementary Road Network of Andalusia. These deficiencies were defined in a checklist and were identified by a road inspection. Decision Trees (DTs), that are a data mining technique that allows the extraction of Decision Rules (DRs), were used. DRs revealed the relationship between road deficiencies and crashes. The application allows two different analyses. A specific analysis of the Complementary Road Network of Andalusia, in which, particular safety problems can be identified, and the location of roads with those problems can be obtained. A more general analysis in which some characteristics related to road safety can be selected in order to know the combination of factors contributing to traffic crashes. Safety problems are based on data from Complementary Road Network of Andalusia but results can be extrapolated to other rural highways in Spain.