Publication

The purpose of this field operational test study is to assess visual attention allocation and brake reactions in response to a brake-capacity forward collision warning (B-FCW), which is designed similarly to all forward collision warnings on the market for trucks. Truck drivers’ reactions immediately after the warning (threat-period) as well as a few seconds after the warning (post-threat-recovery-period) are analyzed, both with and without taking into consideration the predictability of an event and driver distraction. A B-FCW system interface should immediately direct visual attention toward the threat and allow the driver to make a quick decision about whether or not to brake. To investigate eye movement reactions, we analyzed glances 30 s before and 15 s after 60 naturally occurring collision warning events. The B-FCW events were extracted from the Volvo euroFOT database, which contains data from 30 Volvo trucks driving for approximately 40 000 h for four million kilometers. Statistical analyses show that a B-FCW leads to immediate attention allocation toward the roadway and drivers hit the brake. In addition to this intended effect during the threat-period, a rather unexpected effect within the post-threat-recovery- period was discovered in unpredictable events and events with distracted drivers. A few seconds after a warning is issued, eye movements are directed away from the road toward the warning source in the instrument cluster. This potentially indicates that the driver is seeking to understand the circumstances of the warning. Potential reasons for this are discussed: properties relating to the termination of the warning information, the position of the visual and/or audio warning, the conspicuity of the warning, the duration of the warning, and the modality of the warning. The present results are particularly valuable because all on-market collision warning systems in trucks (and almost all in cars) involve visual warnings positioned in the instrument cluster like the one in this study. Acknowledging the fact that human machine interface (HMI)-design is challenging, the conclusions lead the way toward HMI design recommendations for collision warning systems.