Files in this item



application/pdfSAREMI-DISSERTATION-2016.pdf (6MB)
(no description provided)PDF


Title:Participatory sensing fuel-efficient navigation system GreenGPS
Author(s):Saremi, Fatemeh
Director of Research:Abdelzaher, Tarek F.
Doctoral Committee Chair(s):Abdelzaher, Tarek F.
Doctoral Committee Member(s):Vaidya, Nitin; Nahrstedt, Klara; Liu, Xue
Department / Program:Computer Science
Discipline:Computer Science
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):Participatory Sensing
Abstract:The proliferation of smartphones has led to increased interest in mobile participatory sensing. This paradigm enables low cost establishment of a wide range of applications in variety of domains, including environmental monitoring, transportation, safety, healthcare, social networks, urban sensing, etc. This thesis proposes, designs and develops a novel application in this genre, called GreenGPS, which owes its practicality to the widespread usage of smart mobile devices. GreenGPS is a navigation service that finds fuel optimal routes, customized to individual drivers and vehicles, between arbitrary end-points. This thesis studies research challenges revealed in development of GreenGPS on how to build an easy-to-deploy and inexpensive participatory sensing system to support data collection, how to generalize sparse samples of high- dimensional spaces to develop models of complex nonlinear phenomena, how to build a general but personalizable fuel-saving navigation system, how to infer the information on location and type of traffic regulators with low effort and expense, and how to insure reliability of the modeling throughout the lifetime of the service, especially the early deployment stage through which service adoption is sparse and proper modeling facilitates getting the participatory sensing based system off the ground and surviving conditions of sparse deployment. GreenGPS navigation service is offered in both web-based and smartphone application forms. To launch GreenGPS, we deployed a medium scaled vehicular participatory sensing system, consisting of 46 user subjects, collecting over 6700 miles of GPS driving data. To provide a testbed for future transportation fuel saving research, we started to deploy GreenGPS on over 100 vehicles of UIUC Facilities and Services fleet. To give the reader a sense of how effective are route choices provisioned by GreenGPS, it was assessed that compared to alternative fastest and shortest routes provided by traditional navigation tools, green routes are respectively 21.5% and 11.2% more fuel economic. The GreenGPS fuel optimal routes were further compared to Garmin ecoRoutes, a well-known commercial GPS product, and presented 8.4% more fuel savings.
Issue Date:2016-11-23
Rights Information:Copyright 2016 Fatemeh Saremi
Date Available in IDEALS:2017-03-01
Date Deposited:2016-12

This item appears in the following Collection(s)

Item Statistics