5. CONCLUSIONS AND RECOMMENDATIONS

5.1. Summary of Findings

As with any design of a man-machine system, the particular task must be considered in assessing the appropriateness of the fundamental interface devices. Head-mounted displays have been shown to provide an improvement in tracking tasks, but as this study shows, only when tracking is directly coupled with the user's helmet. The task of tracking (regardless of input device) was a suitable task for both display types, because tracking tasks are very common to both output devices in applications and previous studies. The use of external input devices, such as the joystick used in this study, does not appear to impair the subjects' ability to view the tracked environment in either display condition. The converse appears to be true as well; the choice of display did not affect the subjects' performance in tracking with an external input device.

There was no substantial difference in performance between the subjects using the monitor or HMD. Therefore, for a simple task, similar to this experiment the most cost effective visual mode would be preferable. The monitor system provide a satisfactory method of displaying the visual information.

After the experiment, the subject performed one last trial using the opposite visual mode they had in the experiment. They were then asked which mode they preferred. Their response distribution is shown below.

Sixty five percent chose the method they used most during the experiment. This is likely, since they got accustomed to that visual mode. Reasons given by the subjects who disliked the HMD included the difficulty in adjusting the HMD, the HMD felt to confining, and not being able to see the hand controller. The HMD advocates noted the better field of view. Some of the subjects liked having the stereoscopic view of the HMD, even though it wasn't essential. Others found the 3-D distracting. However, all the subjects felt they could perform the task using only the monitor.

5.2. Recommendations

In comparing HMDs with static displays, HMDs are more costly and uncomfortable, and they require more preparation time in using the device. In addition, HMDs restrict the user's movement to a constrained circular area, and focuses the user's attention to the display. Therefore, HMDs are not an effective choice in tasks that require short-term attention, freedom of movement, or the ability to view objects external to the display concurrently. For example, reading electronic mail via HMD would obviously be a poor choice in system design, unless the HMD could offer the user an added benefit over a static display. Since this study shows that HMDs without tracking does not offer any significant benefit over static displays, it is unlikely that HMDs (without tracking) would be heavily considered for common viewing tasks.

However, the unique feature of head-coupled tracking is the added benefit of using HMDs. Determining the use of HMDs in tracking tasks still depends on a number of factors, though. The tracking tasks typically involve a period of time that is uninterrupted by external visual stimuli. Although resolution of the HMD may be sacrificed, HMDs tightly couples head input motion with visual output, which does not deviate far from how people normally track and scan the world. The issues of lag and frame rate must also be considered in the device assessment.

5.3. Areas for Further Study

The choice of visual mode is task dependent. This task used a two dimensional track. The stereoscopic vision of the HMD was not necessary. Another study could look into how much the relative performance of the HMD increase with a three dimensional task. A subject using a monitor would have to adapt to other depth cues to compensate for the lack of stereo vision.

The use of a two-dimensional track has not been common in previous tracking studies; past studies involving tracking have typically used a single-axis path, and allowed the subject to track to a particular target located on the path. The next logical step in furthering this research is to compare the performance of track navigation, using tracks similar to our experiments, between HMD with head tracking and static display with hand-controlled tracking.

Since one of the advantages of a head mounted display is the large field of view. A study looking into field of view of monitors could be performed. How does performance improve as the viewing display becomes larger. This study used 7.5 inch monitors. A further study could see how performance changes using different size monitors. This study showed that it may not be worth using an HMD, but a larger monitor may increase performance.

Continue

Table of Contents