Date Approved

4-26-2018

Embargo Period

4-30-2018

Document Type

Thesis

Degree Name

MS Civil Engineering

Department

Civil and Environmental Engineering

College

Henry M. Rowan College of Engineering

Advisor

Bhavsar, Parth

Committee Member 1

Mehta, Yusuf

Committee Member 2

Chowdhury, Mashrur

Keywords

Autonomous Vehicles, Fault Tree Analysis, System Reliability, Vehicular Components

Subject(s)

Autonomous vehicles; Automobiles--Risk assessment

Disciplines

Civil and Environmental Engineering | Transportation Engineering

Abstract

In 2016, more than 35,000 people died in traffic crashes, and human error was the reason for 94% of these deaths. Researchers and automobile companies are testing autonomous vehicles in mixed traffic streams to eliminate human error by removing the human driver behind the steering wheel. However, recent autonomous vehicle crashes while testing indicate the necessity for a more thorough risk analysis. The objectives of this study were (1) to perform a risk analysis of autonomous vehicles and (2) to evaluate the safety impact of these vehicles in a mixed traffic stream. The overall research was divided into two phases: (1) risk analysis and (2) simulation of autonomous vehicles. Risk analysis of autonomous vehicles was conducted using the fault tree method. Based on failure probabilities of system components, two fault tree models were developed and combined to predict overall system reliability. It was found that an autonomous vehicle system could fail 158 times per one-million miles of travel due to either malfunction in vehicular components or disruption from infrastructure components. The second phase of this research was the simulation of an autonomous vehicle, where change in crash frequency after autonomous vehicle deployment in a mixed traffic stream was assessed. It was found that average travel time could be reduced by about 50%, and 74% of conflicts, i.e., traffic crashes, could be avoided by replacing 90% of the human drivers with autonomous vehicles.

Download

Share

COinS