Development of a Monitoring System for Driver Readiness in Prolonged Automated Driving
-
2024-08-01
-
Details:
-
Creators:
-
Corporate Creators:
-
Corporate Contributors:
-
Subject/TRT Terms:
-
Publication/ Report Number:
-
Resource Type:
-
Geographical Coverage:
-
Edition:Final Report
-
Corporate Publisher:
-
Abstract:Fatal traffic crashes have increased significantly, largely due to human error, which automated vehicle technology aims to reduce. However, challenges such as driver fatigue and the need for quick intervention in case of system failures must be urgently addressed to ensure safety. This study aims to develop a driver fatigue monitoring system to detect different driver fatigue levels during prolonged automated driving. The proposed fatigue monitoring system integrates deep learning, computer vision, and machine learning techniques, leveraging postural and behavioral data. A driving simulator experiment was conducted to collect data on eye aspect ratio (EAR), mouth opening ratio (MOR), percentage of eye closure (PERCLOS), and driver postural information. Computer vision techniques were utilized to extract these features from visual data automatically. The study's key finding is that postural data is the most critical factor in detecting driver fatigue. Among the evaluated machine learning algorithms, the random forest algorithm demonstrated the best performance, achieving an accuracy of 0.97 in detecting driver fatigue. Combining postural data with physical measures such as EAR, MOR, and PERCLOS proved highly effective in accurately identifying driver fatigue levels. This integrated approach offers a promising solution for enhancing the safety and reliability of automated driving systems by effectively monitoring and addressing driver fatigue.
-
Format:
-
Funding:
-
Collection(s):
-
Main Document Checksum:
-
Download URL:
-
File Type: