Effecting the driver’s awareness : Options, bounds, limits
Automotive technology has been developed for decades. During this time the goal of realization of a safe driving stays always the same. Most hybrid electric vehicles (HEVs) provide the “eco” driving mode, which helps to increase the fuel efficiency by changing the vehicle transmission behavior, the acceleration characteristics, the air conditioning system, etc. Besides the alteration of the performance of the vehicle, the behavior of the driver also plays a crucial role in influencing the fuel efficiency. The human driver behaves individually and dynamically. It is necessary to assist the driver in an appropriate manner to behave efficiently. The fully automated vehicle is one of the promising improvements in the automotive technology in the future. As for a transition phase from conventional manual vehicle to fully automated one, vehicles with multi-levels of automated driving are derived. These vehicles consist of several levels of automated driving, which could lead to the problem of mode error/confusion. It is essential to study the limits of the drivers and to assist them to recognize the current situation and the ability of the automated system to ensure a safe takeover. In this thesis, the concept of closing the driver-vehicle-environment loop with a driver-vehicle interface (DVI) is proposed. The bidirectional information flow dis- played on the DVI could help the driver to improve the driving safety and efficiency in different automated driving levels. To realize the increment of fuel efficiency, the DVIs based on the proposed concept displaying the suggested efficiency optimal behavior in different manners as Head-Up Display (HUD) and on instrument cluster are proposed and studied. Two experiments are conducted to compare the efficacy and effect of the DVIs on drivers’ behavior. Results show that the driving efficiency could be improved by showing the suggestion. The most efficient drive requires the most visual demand. By considering the distracting factors from the environment into the suggested optimal behavior, the cognitive workload could be reduced. For the vehicles with multi-levels of automated driving, novel interactive DVIs are proposed to reduce the mode error/confusion. The results from previously performed two experiments are applied to the lower automated levels. A third experiment is executed to evaluate the efficacy and acceptance of such DVIs as well as to study the takeover behavior of the drivers in various critical situations. The proposed DVIs are positively confirmed by the participants. The obtained results show that the takeover time in the first drive is significantly higher than the following ones. This means that using only the average takeover time to design the integration of human as a fallback layer of automated vehicles is not suitable. It is concluded that the takeover time also depends on the individual driver, the driving velocity, the difficulty of the critical situation, etc. The proposed interactive DVIs are adapted based on the conclusions from the conducted experiments. They show the information from different automated driving levels systematically using the HUD, the instrument cluster, the rear view mirrors, and the touch display. The functionalities are dynamic to the active driving mode. The warnings and suggestions to the driver are displayed dynamically based on the tracked glance position. New research questions are posed regarding safe takeover, comprehensive understanding of the TOR as well as qualitative and quantitative relations between the takeover and the activity of driver in non-driving-related task (NDRT). Based on the performed experiments and obtained results, the discussion about the design goals and principles as well as requirements for adaptive DVIs for future development are concluded in this thesis. The future DVIs should be able to show the situation-dependent information appropriately, to learn, or to adapt to the individual behavior of the driver, as well as to combine the driving efficiency, the allowable situation, and the intention of the driver.