Degree Type

Creative Component

Semester of Graduation

Spring 2019

Department

Computer Science

First Major Professor

Simanta Mitra

Second Major Professor

Gurpur Prabhu

Degree(s)

Master of Science (MS)

Major(s)

Computer Science

Abstract

Over the last decade, it is estimated that 25% of the congestion on the US roads is due to traffic incidents such as crash, overturned trucks or stalled vehicles. Currently available software based intelligent transportation systems does not provide comprehensive decision support to minimize the impact of traffic incidents and do not detect the incidents on time. The aim of the TIMELI project is to develop a robust and fast incident detection on the road. Lambda architecture is used to design the architecture of TIMELI project. Different technologies are explored to finalize the best design to accomodate the given functional and non-functional requirements. In this incident detection method, real time data of speed of each segment on the road is recorded every minute. This data is used to detect any congestion or incident anomalies and alert the TIM(Traffic Management Operator) immediately. Also, it can automatically close the incident. The algorithm also captures the start of the incident as well as time it ended. The incidents are stored for data analytics and incident validation and performance.

Comments

This paper using traffic speed data and detects incidents in real-time. Also describes in detail architecture of real time processing in TIMELI project.

Copyright Owner

Raviteja Raja Byna

File Format

application/pdf

Share

COinS