Thank you, Mr. Chairperson. Good morning, everyone. I'm pleased to have this opportunity to talk about our Drowsiness Warning System.

At Nissan, we have been researching ways of detecting drowsy driving for about 20 years now.

Fifteen years ago, we introduced the Safe Driving Advisor, which detected drowsiness on the basis of certain steering patterns, and issued an audible warning to alert the driver.

Test results showed that a state of drowsiness could be detected on the basis of alpha brain waves. And we verified that the occurrence of certain distinctive steering patterns coincided with the appearance of certain alpha wave patterns.

After that, we researched more accurate ways of detecting drowsiness. Techniques for detecting drowsiness can be broadly divided into five categories. They have the general characteristics shown in this table.

In this research, we focused on a system for detecting changes in the degree of openness of the driver's eyes. This approach falls under the category of detection of physical changes in physiological phenomena. This particular method was selected because a practical drowsiness detection system must provide a high level of detection accuracy. Moreover, the system should be able to detect drowsiness by means of a noncontact technique.

This is the configuration of the drowsiness warning system. A small CCD camera positioned in front of the driver takes images of the driver's face. The facial image data are converted to binary images one frame at a time and sent to the frame memory of the image processor. A personal computer connected to the image processor controls the image processing procedure and judges the processed results. An infrared lamp is provided in the instrument panel to facilitate the recording of facial images during nighttime driving.

As the level of alertness drops, rapid blinking gives way to the appearance of long intervals when the eyes are closed, which provides a basis for detecting drowsiness. The method is to count the number of times the eyes close within a specified interval. The count begins with the second consecutive closure. This is done to avoid including instances of eye closure due to blinking. The numbers here in the middle of the interval for judging the alertness level indicate the eye closure count. In this example, the system judged that the eyes closed four times.

Having found that neither a scent nor a buzzer alone achieved a sufficiently long refreshing effect, we investigated the effect obtained by providing a combination of stimuli. With the combined stimuli, a buzzer was first sounded to give the subject an immediate refreshing effect and a scent was then released once the subject's alertness had risen to a certain level. As the results shown here indicate, the combination of stimuli heightened the refreshing effect to some extent, compared with the presentation of a single stimulus alone.

(May we have the last slide, please.)

Today, I have described the operation of our Drowsiness Warning System, which detects drowsiness using a noncontact method to monitor the driver's alertness level. I have also presented test results that indicate a combination of audible and scent stimuli is more effective in refreshing the driver than either stimulus alone.

Thank you very much.