Konzeptionierung, Auslegung und Umsetzung von Assistenzfunktionen für die Übergabe der Fahraufgabe aus hochautomatisiertem Fahrbetrieb

Abstract

The ongoing development of driver assistance systems leads to a continuous increase in the degree of automation of today's vehicles.Whilst latest systems already carry out lateral or longitudinal guidance to a big amount, the combination of these functions will lead to the first automated car.The fact that this prediction is not fictive is known since at least the Darpa Urban Challenge in 2007.In this race robot cars competed against each other in an urban environment.However, until this point can be reached many challenges have to be faced.Next to unanswered questions dealing with responsibilities and licensing, even challenges in the field of technology have to be taken.Therefore the introduction of automated cars into the market, is expected to be carried out step by step.Whilst nowadays already partly automatic functions are being used in series cars, the next step will be automatic driving for specified time.In that intermediate step towards full automation, cars will handle the whole driving task for defined time and or situations.As soon as system boundaries are reached, the driver will be asked to take over control again.This handover of control from highly automated driving can be a hard challenge for the driver.First of all a totally distracted driver should be detached from his active task.After that he should recognize the traffic situation as fast as possible, followed by actions to take over control again.The available time for these actions depends on the one hand on the sensors of the car and on the other on the driver.This thesis starts with the basis of theoretical models of the driving task, as well as human driving behavior from which the challenges to face are being derived.On this basis strategies are designed and assembled which support the driver in these situations.The support strategies are designed according to the three phases that have to be carried out in these viii situations: take over signal, recognition of the traffic situation and lateral guidance assistance after the actual takeover.The support strategies are tested in three studies in a driving simulator with naive test subjects.By the use of a driving simulator the advantages of early and close to reality testing are taken.By using a takeover signal and thus deactivating the side task a low takeover time was reached.The support in the cognitive phase is carried out by the presentation of situation describing images.The subjective positive aspects of this cognitive support face the effect of longer takeover times, which are caused by the visual focusing of the images.After a successful takeover drivers are supported by the lateral guidance system, which especially for the first seconds leads to a significant influence on lateral guidance performance.