3rd International Conference on ESAR
Filtern
Schlagworte
- Active safety system (3) (entfernen)
Methods for analyzing the efficiency of primary safety measures based on real life accident data
(2009)
Primary safety measures are designed to help to avoid accidents or, if this is not possible, to stabilize respectively reduce the dynamics of the vehicle to such an extent that the secondary safety measures are able to act as good as possible. The efficiency of a primary safety measure is a criterion for the effectiveness, with which a system of primary safety succeeds in avoiding or mitigation the severity of accidents within its range of operation and in interactionwith driver and vehicle. Based on Daimler-´s philosophy of the "Real Life Safety" the reflection of the real world accidents in the systems range of operation is both starting point as well as benchmark for its optimization. This paper deals with the methodology to perform assessments of statistical representative efficiency of primary safety measures. To be able to carry out an investigation concerning the efficiency of a primary safety measure in a transparent and comparable way basic definitions and systematics were introduced. Based on these definitions different systematic methods for estimating efficiency were discussed and related to each other. The paper is completed by presenting an example for estimating the efficiency of actual "single" and "multi" connected primary safety systems.
Over the last decades the number of traffic accident fatalities on German roads decreased by 77% down to 4968 in the year 2007. This positive development is due to optimisations of vehicle safety, roads and infrastructure and medical rescue issues. Up to now mostly the optimisations of secondary safety measures lead to this effect on vehicle safety. Since some years more and more driver assistance systems are available and lead to a further reduction of all accidents. These new systems are often comfort systems and have not primarily been developed to increase vehicle safety. In contrast to secondary safety systems primary safety systems are able to mitigate and avoid accidents. So in the future it is important to estimate the benefit of these systems in reducing accident numbers as well. Current benefit estimation methods mostly focus on a single system only and not on the combination of systems. In this paper a new method for a multivariate benefit estimation based on real accident data is developed. The paper describes the basic method to estimate the benefit of primary and secondary safety systems in combination. With the presented method the benefit will not be overestimated as it would be by a simple addition of the benefits of single systems. The model will be validated by a multivariate prospective benefit estimation of different vehicle safety systems in comparison to single benefit estimations of the same systems. For this the German In-Depth Accident Database is used. The results show the importance to implement the interactions of safety systems in the estimation process and rate the overestimation by a simple addition of the single system benefits. The validation includes primary and secondary safety systems in combination. The validation is done using more than 3500 real accidents which were initiated by cars. This sample out of the GIDAS database is representative for the current accident situation in Germany. The paper shows the necessity of a multivariate estimation of the benefit for existing and future safety systems.
It has been pointed that most of the accidents on the roads are caused by driver faults, inattention and low performance. Therefore, future active safety systems are required to be aware of the driver status to be able to have preventative features. This probe study gives a system structure depending on multi-channel signal processing for three modules: Driver Identification, Route Recognition and Distraction Detection. The novelty lies in personalizing the route recognition and distraction detection systems according to particular driver with the help of driver identification system. The driver ID system also uses multiple modalities to verify the identity of the driver; therefore it can be applied to future smart cars working as car-keys. All the modules are tested using a separate data batch from the training sets using eight drivers" multi-channel driving signals, video and audio. The system was able to identify the driver with 100% accuracy using speech signals of length 30 sec or more and a frontal face image. After identifying the driver, the maneuver/ route recognition was achieved with 100% accuracy and the distraction detection had 72% accuracy in worst case. In overall, system is able to identify the driver, recognize the maneuver being performed at a particular time and able to detect driver distraction with reasonable accuracy.