Home
Resource
Safe driving with mobile devices and wearables
PDF
PDF format is widely accepted and good for printing.
Plug-in required
PDF-1(8.50 MB)
Citation & Export
View Usage Statistics
Staff View

Citation & Export
Hide

Simple citation

Karatas, Cagdas. Safe driving with mobile devices and wearables. Retrieved from https://doi.org/doi:10.7282/T3GH9N54

Export

Click here for information about Citation Management Tools at Rutgers.

Statistics
Hide

Description

TitleSafe driving with mobile devices and wearables
NameKaratas, Cagdas (author); Gruteser, Marco (chair); Rutgers University; School of Graduate Studies
Date Created2018
Other Date2018-01 (degree)
SubjectElectrical and Computer Engineering, Wearable technology, Automobiles--Safety measures, Distracted driving
Extent1 online resource (xiv, 108 p. : ill.)
DescriptionDriver errors due to distracted driving and inadequate awareness of surroundings is an increasing concern which has led to national attention. With increasing amount of information available through the ecosystem of in-vehicle devices like smartphones, wearables, and mobile OS for cars, opportunities to prevent accidents with safety services are now more available than ever before. The safety services are expected to be aware of the driver’s actions and car’s context, and prevent unsafe driving. Mobile and wearable safety apps differ from conventional built-in automotive safety systems in that they promise low-cost designs that reach a much larger population. On the other hand, detecting the driver’s actions and the car’s context from mobile sensors is a challenging task due to the dynamic nature of the car and the sensors. To assist in safer driving, we designed and evaluated fundamental solutions that can be used to detect and prevent driver errors. First, we propose methods to monitor the vehicle, driver’s steering, and attention to detect driver errors. The preventive methods are auxiliary services which decrease the need for distracting interactions with the vehicle and phone. In order to detect driver errors, mobile devices and wearables such as wrist-worn devices and head-mounted devices can be used. The mobile device can be used to monitor the vehicle’s movements while the wearables can be used to monitor the driver’s head and arm movements. We are particularly interested in the driver’s hand and head ii movements since these type of movements shows the user’s attention and the driving interactions with the vehicle, particularly with the steering wheel. Additionally, the detection techniques proposed in this thesis can be also be further utilized to prevent errors by warning the drivers about dangerous conditions. Preventive techniques propose a mechanism for convenient interaction between the user and the environment. Examples of preventive techniques include easy-to use customizable input interfaces as well as management of notifications to appropriate communication channels and scheduling them to the most convenient times, e.g. when the vehicle stops at a red traffic light. Through various real-driving scenarios, we show that our approach can detect the wrist-worn device user as driver correctly 98.9% of times and achieve hand on steering wheel detection with a true positive rate around 99% and provide warning of unsafe driving when a driver’s hand is off the steering wheel with a true negative rate above 80%. Additionally, the system can achieve accurate steering angle estimation with errors less than 3.4 degrees to facilitate applications such as curve speed warning and understeer/oversteer detection. In the second part of our system, we introduced a novel method to estimate sensor orientation and the vehicle’s heading from only a single inertial sensor in a moving vehicle. The method was able to estimate sensor orientation with a mean error of 5.61o for yaw angle and a 3.73o for pitch angle while the vehicle is driven in controlled environment and without restrictions. We believe this method can be especially suitable for head tracking applications where the sensor’s translational motion is limited. Finally, we proposed a framework that enables users to create customizable printable paper button interfaces that might be used to create shortcuts and reduce mobile-device usage related distractions. Our experiments indicate that the sensor achieves touch detection accuracy over 99% with up to ten different touch points and over 90% with 15 different touch points.
NotePh.D.
NoteIncludes bibliographical references
Noteby Cagdas Karatas
Genretheses, ETD doctoral
Persistent URLhttps://doi.org/doi:10.7282/T3GH9N54
Languageeng
CollectionSchool of Graduate Studies Electronic Theses and Dissertations
Organization NameRutgers, The State University of New Jersey
RightsThe author owns the copyright to this work.
Version 8.5.5
Rutgers University Libraries - Copyright ©2024