Design of a 3D interactive simulator for driver behavior analysis

Abstract

Diagnosing the causes of road accidents and the development of effective countermeasures to reduce accident rates is of key importance in road safety. Human error is one of the principal influencing factors that lead to road accidents, and is attributed to increased mental workload induced by distractions. Workload, however, is characterized by intrinsic properties that are difficult to observe. Hence, phenotype behaviors, such as lane deviations, could act as good predictors of driver workload. Driving simulators emerged as a promising technology for the analysis of driving conditions and road users' behavior in an attempt to tackle the problem of road accidents. However, the cost of designing or owning a simulator to conduct a safety analysis is prohibitive for many government agencies. The work presented herein demonstrates the design and development of a driving simulator, using a 3D game engine that aims to contribute towards evaluating black spots in road networks by promoting rapid design of realistic models and facilitating the specification of test scenarios. The developed simulator was employed to conduct a set of preliminary experiments that analyzed driving behaviors of local road users for a chosen black spot in a road network in Limassol-Cyprus. Data collected from the experiments are analyzed, results are presented and conclusions are drawn.