Impact of the geometric field of view on drivers’ speed perception and lateral position in driving simulators

Abstract

Driving simulators have become an effective tool in road safety research. In recent years, the validity of simulators raised debates concerning the extant to which driving in the simulator resembles driving in the reality. Different types of driving simulators with different characteristics have been developed to study driver behavior, however, the fidelity and reliability of such systems are questionable if no proper validation is conducted. Regarding the visual aspect, the fidelity of the simulators can be assessed based on the field of view of the simulator screens. Drivers' speed perception and lateral position were compared for two different geometric field of view (GFOV) angles (i.e., 60 and 135 degrees). Results from the ANOVA tests showed that drivers highly underestimate their driving speed while driving for the condition with 60 degrees of GFOV compared to the condition with 135 degrees of GFOV. Furthermore, drivers drove closer to the real-world situations in the condition with 135 degree of GFOV compared to the condition with 60 degree. Results of this study suggest that, using incorrect GFOV for any simulator would generate biased results in speed and lateral position. Therefore, a proper calibration criterion of the GFOV for the simulators is essential. This study recommends using a scale factor (GFOV/FOV) of 1.00 for virtual environment offered by the simulation scenarios such as GFOV of 135 degree for simulators having three screens with 135 degree of field of view (FOV). Abstract Driving simulators have become an effective tool in road safety research. In recent years, the validity of simulators raised debates concerning the extant to which driving in the simulator resembles driving in the reality. Different types of driving simulators with different characteristics have been developed to study driver behavior, however, the fidelity and reliability of such systems are questionable if no proper validation is conducted. Regarding the visual aspect, the fidelity of the simulators can be assessed based on the field of view of the simulator screens. Drivers' speed perception and lateral position were compared for two different geometric field of view (GFOV) angles (i.e., 60 and 135 degrees). Results from the ANOVA tests showed that drivers highly underestimate their driving speed while driving for the condition with 60 degrees of GFOV compared to the condition with 135 degrees of GFOV. Furthermore, drivers drove closer to the real-world situations in the condition with 135 degree of GFOV compared to the condition with 60 degree. Results of this study suggest that, using incorrect GFOV for any simulator would generate biased results in speed and lateral position. Therefore, a proper calibration criterion of the GFOV for the simulators is essential. This study recommends using a scale factor (GFOV/FOV) of 1.00 for virtual environment offered by the simulation scenarios such as GFOV of 135 degree for simulators having three screens with 135 degree of field of view (FOV).