Abstract
Keywords - brake assist system, driver adaptive warning, onboard analysis of driver behaviour, driver type, active safety
The various design aspects of advanced driver assistance systems are challenging: next to the main development objective, to enhance the safety by foresighted driver support, it is necessary to observe legal regulations and to ensure the driver´s acceptance. Once the first technical hurdles, e.g. detecting and classifying obstacles or analyzing the vehicle environment, were taken, systems for a wide variety of situations will be available in the coming years. But though technical requirements are fulfilled, intra- and inter-individual needs and expectations of drivers that have to be taken into account will limit the possibilities of numerous active safety systems: the kind of human-machine-interface and thresholds for warning and intervention can only be adapted to the demands of an average group of drivers. Both old or anxious drivers, who may have a need for more support, and sportive ones, who are quick on regarding intervening systems as unnecessary, cannot be supported ideally. Therefore, driver adaptive systems with driver-type identification are required to use the full capacity of assistance systems: they ideally can provide fully customized support and allow to increase the safety benefit by meeting the driver´s individuality.
In this paper, a possible implementation of a driver adaptive strategy is presented, based on currently available brake assist systems (BAS). Previous studies of the safety benefit of advanced BAS have shown that strategies with driver warnings and autonomous partial braking have a notable potential for collision avoidance: the injury risks can be reduced significantly. However, these strategies have a great impact on the decision making-authority of the driver and his feel of comfort. These effects can be compensated by a driver-adaptive warning system which is able to classify different types of drivers on the basis of measurable variables such as braking deceleration, velocity and safe distance. Significant correlations between driver and dynamic vehicle behaviour were found in different studies. A safety benefit using an adaptive algorithm in assistance systems can be expected in follow scenarios as well as in situations where a vehicle approaches an obstacle or the tail end of a traffic jam. Moreover, an advanced BAS would be able to display only relevant information e.g. about the needed safe distance and to start a partial braking faster than conventional systems, if the driver doesn´t react as required.
But not only assistance systems can benefit from driver classifications: the identification of the individual driver connected with information about his expectations and needs makes it also possible to adapt the design of e.g. automatic gearboxes or fuel consumption strategies.