An object oriented Bayesian network approach for unsafe driving maneuvers prevention system

摘要

As the main contributor to the traffic accidents, unsafe driving maneuvers have taken attentions from automobile industries. Although driving feedback systems have been developed in effort of dangerous driving reduction, it lacks of drivers awareness development. Therefore, those systems are not preventive in nature. To cover this weakness, this paper presents an approach to develop drivers awareness to prevent dangerous driving maneuvers. The approach uses Object-Oriented Bayesian Network to model hazardous situations. The result of the model can truthfully reflect a driving environment based upon situation analysis, data generated from sensors, and maneuvers detectors. In addition, it also alerts drivers when a driving situation that has high probability to cause unsafe maneuver to be detected. This model then is used to design a system, which can raise drivers awareness and prevent unsafe driving maneuvers.