Abstract:The primary objective of in-vehicle route guidance systems is to reduce congestion by influencing drivers to select routes which will probably be beneficial to them individually and to the system as a whole. It is obvious that the driver must perceive, recognize the guidance information, make a decision and execute his/her action appropriately. In this regard there is a concern that the presence of the in-vehicle route guidance system will take away the driver's attention from the primary driving task. The primary objective of this research project is to study how in-vehicle route guidance system attributes, driver characteristics and traffic conditions affect driving performance.

A series of realistic experiments were conducted using a driving simulator developed by the Hughes Aircraft Corporation, in order to examine the implications of different types of route guidance systems. The focus of the experiments were on en route guidance over predescribed and planned routes. Drivers were asked to follow a predetermined route to the destination. Four types of route guidance systems were tested. They are: (i) Paper Map, (ii) Heads Down Electronic Map, (iii) Heads Up Display (HUD) in combination with Electronic Map, and (iv) Voice Guidance in combination with Electronic Map.

The experiment was designed so that all subjects were tested using all four route guidance systems, in order to analyze the effect of route guidance type as a within subject experiment. The order of presentation of the systems and their assignment to driving scenarios were randomized using a latin square design. Subjects were recruited by a market research firm under instructions from UC Davis Researchers. Data were collected for a total of 18 subjects, 9 male and 9 female. The male and female groups were further divided on the basis of driving experience:

• Low experience for driving less than 12,000 miles in the previous year.
• High experience for driving greater than 15,000 miles.

An inobtrusive eye tracker was used to monitor eye fixations for the 3 electronic systems. User perceptions and preferences for the devices and the subjects subjective assessment of workload were measured using a variety of measuring scales. A number of performance measures were automatically collected by the simulator including: (i) Subject vehicle speed; (ii) Headway (time from front bumper of lead vehicle to front bumper of subject vehicle); (iii) Lateral position; and, (iv) Reaction times to external events. Headway, lateral position and speed were measured every 0.1 second. Reaction times were measured whenever the subject had to react to an external event to avoid a collision. External events included: pedestrians crossing in front of the subject, left turning vehicles, crossing vehicles and obstacles. In addition, the reaction times to changes in traffic signal indication from green to amber were also collected. The NASA TLX method was used to measure subjective workload immediately after completing the driving trial with each of the four route guidance systems. Driver preferences of each system were also measured at this time using 5 dimensions: ease of use; clarity of information; quantity of information; preparation for turns; and, levels of distraction.

The following performance measures were considered for the analysis: reaction times, workload ratings, user perceptions, percentage of dwelling time on the road (from the eye tracker) and the number of navigation errors. For reaction times, ANOVA models were developed. For workload and percentage of dwelling time on the road, linear regression models were developed. For the number of errors, a logistic regression model was developed with the dependent variable being the probability of the occurrence of an navigation error. Independent variables were type of route guidance system, subject category (sex and driving experience) and order effects. Order effects were included in the models to test for the presence (if any) of changes in driving performance that may be attributed to the order in which the devices were presented to the subject.

Major findings of the study may be summarized as follows:

• (1) Subjective workload, user perceptions, eye tracker dwelling times and number of errors all indicated that the voice guidance/electronic map combination performed the best, and the paper map to be the worst. The electronic map was found to be the second best, closely followed by the HUD electronic map. Somewhat surprisingly, the heads up HUD/electronic map combination performed worse than the electronic map in the case of workload.
• (2) The reaction time modeling yielded slightly different device performance depending on the event being reacted to. The heads up display/electronic map combination performed much better in comparison to its performance in the other performance measures with voice/electronic map also doing well. The paper map again consistently performed the worst.
• (3) Driving performance did vary with gender and experience. Not surprisingly, drivers with higher experience performed better than drivers with lower experience. This effect was more prominent among females than males.
• (4) In general, there seemed to be a tendency for the reaction times to increase in the last period (route guidance technique) for each of the subjects. This could have been due to a combination of fatigue and anticipation of the end of the experiment.