Due to the escalating usage of cellphone and social networking, distracted driving is and will remain as one of the most serious problems faced by Departments of Transportation (DOTs) and law enforcement agencies. Under the aim of in-depth investigation of distracted driving crashes in Louisiana, the specific objectives of this study are: (1) reviewing the crash reports for the quality of distracted driving crash reporting, (2) analyzing distracted driving-related crashes through regression model and data mining algorithm to link the severity of distracted driving crashes with the contributing factors collected in crash data, (3) investigating the observable characteristics of distracted driving roadside and video survey, and (4) recommending the countermeasures utilizing the analysis results and reviews. About 60,000 crashes from ten-year crash data, three types of distracted driving related crashes are modeled: Fatal (K) and Severe (A) Injury;Moderate (B) and Complaint (C) Injury; and Property-damage only (PDO). One statistical method was used for prediction, multinomial logistic regression, and one data mining algorithms was used, random forest. Higher speed limit, curved road, head-on crashes were identified among the key factors. Data mining algorithms performed better in prediction compared to the multinomial logistic regression when sensitivity and specificity were used to compare the predicted results. Fisher’s exact tests of roadside manual observation data shows that gender has no significant influence in cellphone distraction (regardless of distraction type), however age can be influential and associated with driver distraction. Association rule mining of observation data shows that the most predominant type of cellphone use is manipulating mainly occurs at intersections, whereas talking is more associated with segments. In-vehicle video data were coded by the software FaceReader, which captures facial expressions of drivers while driving. Initial results do suggest valence in emotion can be attributed to timing before, during, and after cellphone calls and texting. Physical countermeasure development towards reducing the distraction-related crash severity should be targeted at preventing lane departure crashes. Physical countermeasure development towards reducing the distraction-related crash severity should be targeted at preventing lane departure crashes. Strict enforcement of texting ban with awareness campaign are also expected to prevent distracted driving.


    Zugriff

    Zugriff über TIB

    Verfügbarkeit in meiner Bibliothek prüfen


    Exportieren, teilen und zitieren



    DETECTION OF DISTRACTED DRIVERS

    ESNA ASHARI ESFAHANI ALIREZA / MUDALIGE UPALI P | Europäisches Patentamt | 2024

    Freier Zugriff

    Drivers with child passengers: distracted but cautious?

    Maasalo, Ida / Lehtonen, Esko / Summala, Heikki | Elsevier | 2019


    Detecting Distracted Drivers Using Convolutional Neural Networks

    Tarabin, Mohamad / Alketbi, Mariam Mohamed / Alfalasi, Hamad Rashed et al. | IEEE | 2023


    Distracted driving in elderly and middle-aged drivers

    Thompson, Kelsey R. / Johnson, Amy M. / Emerson, Jamie L. et al. | Elsevier | 2011


    Adaptive warnings and emergency braking for distracted drivers

    DESHPANDE ANUP S | Europäisches Patentamt | 2022

    Freier Zugriff