Special | Report
Augmented Reality-Based Display Heightens Driver’s Safety
After optimizing specifications, Maxell has verified that the installation of AR-HUD has reduced gaze point and focusing time of the driver compared with conventional Head-up displays.
I

n recent years, legislations related to safe driving assistance systems have advanced on a global scale, and various measures have been taken mainly in countries in Europe and in the United States. Against this backdrop, Augmented Reality Head-Up Display (AR-HUD), which displays driving assistance information, like speed and alert information, superimposed on the windshield through which the driver can see the actual scene, is now attracting attention. As AR-HUD can provide driving assistance information while minimizing the movement of the driver’s gaze point while driving, the demand for AR-HUD is expected to grow as a measure to prevent the driver from taking eyes off the road ahead while driving.

Seek Optimal Specifications for AR-HUD
Maxell, Ltd. has been conducting extensive tests by driving a car equipped with AR-HUD in urban areas and on expressways during daytime and nighttime in both fine and rainy weather, for the purpose of optimizing the display of driving assistance information, such as display position, size, time, and method. In particular, as the display position of information is crucial from the viewpoint of safety, the movement of the driver’s gaze point is quantified under various conditions using an actual vehicle.

Measurement of gaze point during driving
Fig. 1: Measurement of gaze point during driving (Definition of viewing angle)
Quantification of movement of driver’s gaze point during driving
Analysis of data of the gaze-point movement during driving obtained from test drivers with different ages, gender, and driving experiences was conducted. Fig. 1 shows the scene viewed from the driver’s seat when driving at 30km/h in the city during daytime in fine weather. In order to quantify the gaze-point movement in the vertical direction, which depends on the scenery as seen by the driver, measurement of the vertical movement of the gaze point with respect to the horizon as the reference (viewing angle of 0°) together with the gaze-point distance* while driving, was done. The correlation chart of these data is shown in Fig. 2.
Measured driver's gaze-point movement
Fig. 2: Measured driver’s gaze-point movement
Examination of results
Based on the results on city driving, it was found that the driver pays attention to the speed meter and the landscape with a distance of 40m ahead or closer. In addition, it was also confirmed that the driver looks at the speed meter with a viewing angle of −20°, occupying about 20 percent of his driving time, thus showing a significant shift in the line of sight.

Data during high-speed driving were also analyzed. It was confirmed that the driver’s gaze-point distance is at 16m at low speed, 100m at high speed, and the vertical Field of View (hereinafter referred to as “FOV”) is 4°.

On the other hand, it is important to display information within a range of about 18m at 100m ahead, taking into account the display of possible deviation from the driving lane, and the moving direction of the vehicle, and the road width (based on the standard in Japan). The corresponding horizontal FOV was found to be 10°.

Safety Improvement by Installing AR-HUD
Based on the above results, a prototype of AR-HUD with FOV (vertical 4°, horizontal 10°) and virtual image distance of 20m has been developed, and used for the verification of safety support effects of the installed AR-HUD. In the following, the evaluation conducted on expressways are described. The safety support effect is quantified under three conditions: (i) HUD not installed, (ii) combiner type HUD installed, and (iii) AR-HUD installed.

Quantification of driver’s gaze-point movement and focusing (focus adjustment) time
The range of gaze-point movement and focusing time after each movement were measured for drivers driving on an expressway during the daytime for the purpose of verification of the safety support effect.

Examination of results
Fig. 3 shows a picture, taken at the driver’s seat, of the area as seen by the driver when driving on an expressway. It was found that without an HUD, the driver sees a wide range including the speed meter. When an HUD is installed, on the other hand, the range of gaze-point movement becomes narrow. This shows that a high safety support is achieved because the driver can watch only the front (some distance away).

Driver's gaze point when driving
Fig. 3: Driver’s gaze point when driving on an expressway
Next, comparison and evaluation were made for the time required for the driver’s gaze point to move from the driving information displayed in the car to the front scene and the focusing time. When an HUD was not installed, the time required for the gaze-point movement from a speed meter was used for evaluation. As shown in Table 1, when the gaze point moves from the front scene to the AR-HUD display image, a total of 0.08 sec on average of the gaze-point moving time and focusing time has been achieved. This confirmed significant improvement in comparison to the cases without an HUD and with a combiner HUD.

The verification mentioned above proved that safer and more comfortable driving is supported by installing AR-HUD, which can display driving assistance information at a long distance (virtual image distance of 20m).

Table 1: Measured focusing time
Measured focusing time
HUD Miniaturization Technology
An HUD is mounted on the space between the instrument panel and the windshield in front of the driver (Fig. 4). As air conditioner ducts and other products are also installed here, miniaturization is an important issue.
Structure of prototype HUD
Fig. 4: Structure of prototype HUD
Maxell has developed a backlight unit with high light utilization, in which the HUD optical path length is significantly shortened by using a free-form curved surface optical technology and special optical elements. This has realized miniaturization, reduced power consumption, and large screen display (equivalent to 148 inches).

Summary
AR-HUD features an augmented reality function that overlays landscapes and images. As mentioned above, it has significantly reduced gaze-point movement and focusing time in comparison with conventional HUD, and therefore supports safer and more comfortable driving. In the future, the company will further promote miniaturization of the equipment in order to increase the percentage of vehicles equipped with AR-HUD.

Note:
* Gaze-point distance: Distance from driver to gaze point

About This Article:


The author, Akio Misawa, is Manager at the AIS Design Department, Optronics Division, Maxell, Ltd.