Sleep related crashes have received increasing attention during the last decade. One possible countermeasure is driver support systems that predicts driver sleepiness and warn the sleepy driver. Of natural reasons it is very difficult to evaluate the effects of such a system. The aim of this study was to test a method for evaluation of a warning system for sleepy drivers, but also to evaluate the effects related to the use of a specific warning system, consisting of a warning strategy with HMI devices as sound, spoken messages, a vibrations unit and a confirmation button. The experiment took place on a closed racing circuit at night. The circuit is in total 4 kilometres long and 40 subjects participated. The design was carried out as a between subject design. The drivers were divided into four different groups (10 subjects in each): (1) Baseline without any feedback or warning; (2) Warning system with an early trigger; (3) Warning system with a late trigger; and (4) Feedback and warning with a late trigger. The start of the warning system (trigg) was done manually by the test leader. The decision was taken with support from the ORS level. The ORS is an observation rating method, developed for this experiment and is based on observations of the driver and driver behaviour. The results are divided into sections in order to describe if the warning system will help the drivers to be more aware of the sensation of sleepiness (unconscious or conscious) and if the system will make them do more countermeasure (not effective or effective) compared to those without the warning. Two additional section focus on the evaluation of the HMI and on the influence on the driving behaviour caused by the warning system. The result showed that most of the participants (23 out of 27 that experienced warnings) reported that the warning made them more awake. All drivers were not however so sure that it had helped them in any way; 4 drivers stated that it had not helped them and 6 drivers did not know; 15 drivers thought it had helped them. Among those receiving warnings 15 drivers thought the warning had come at the right time, 3 thought it was too late and 8 did not know. There were no observable significant differences in driving behaviour between the groups with a warning system and the group without warning system. There was no observable difference between the groups that received the warning in an early phase, late phase or late in combination with a feedback system. The sound was seen as most disturbing and most effective. The vibration at the belt was seen as least disturbing but also as least effective. This seems to be highly related to the level of the sound and the amplitude of the vibrations. (A). For the covering abstract of the conference see E216632.
Samenvatting