A study to measure the accuracy of speed data reported by floating car data in rural areas

Simple item page
dc.contributor.advisorBruwer, Meganen_ZA
dc.contributor.authorKiautha, Antony Mugambien_ZA
dc.contributor.otherStellenbosch University. Faculty of Engineering. Dept. of Civil Engineering.en_ZA
dc.date.accessioned2023-02-23T13:08:19Zen_ZA
dc.date.accessioned2023-05-18T07:15:00Zen_ZA
dc.date.available2023-02-23T13:08:19Zen_ZA
dc.date.available2023-05-18T07:15:00Zen_ZA
dc.date.issued2023-03en_ZA
dc.descriptionThesis (MEng)--Stellenbosch University, 2023.en_ZA
dc.description.abstractENGLISH ABSTRACT: Intelligent Transportation Systems (ITS) technologies are becoming routine worldwide, offering newer methods of obtaining traffic data. Recently, on-board vehicle navigation devices have offered Global Positioning System (GPS) data spanning an entire road network referred to as Floating Car Data (FCD). The accuracy of FCD is critical to evaluate traffic information and inform stakeholders involved in transportation planning in different regions. This study was conducted to measure the accuracy of speed data reported by commercial FCD in rural areas. The study compared benchmark speed data and TomTom historical speeds on 12 rural routes and 7 urban routes in the Western Cape in South Africa. The benchmark data was provided by the Western Cape Department of Transport, collected through loop inductance. Data collected was analysed for weekdays (Monday to Friday) in the month of February 2019 and aggregated over 1-hour time intervals. This was done to give a true representation of traffic patterns and analyse traffic patterns before the COVID-19 pandemic. To evaluate accuracy, probe penetration rate, signed error bias (SEB), average absolute (AASE) and signed error were determined. The minimum probe penetration rate considered for accuracy measurements was 4% and the maximum allowable errors for SEB, AASE and signed error were ±7.5 km/hr, ±10 km/hr and ±10 % respectively. The SEB and AASE accuracy measures showed that FCD reported speeds were consistently lower in the urban areas while some routes in the rural areas recorded higher FCD speed estimate than the benchmark data. Probe penetration rates for both urban and rural areas indicated that a high probe penetration rate is not directly proportional to a high level of speed data accuracy. This study is relevant because evaluation of commercial FCD accuracy has not been examined in rural areas and it is an informative research to assist in improving the transportation industry across the globe and especially in developing countries.en_ZA
dc.description.abstractAFRIKAANS OPSOMMING: Intelligente vervoerstelsels (ITS)-tegnologieë word wêreldwyd benut, en bied nuwe metodes aan om verkeersdata te bekom. Voertuignavigasietoestelle het redelik onlangs Global Positioning System (GPS) data begin verskaf wat oor die hele padnetwerk beskikbaar is: Floating Car Data (FCD). Die akkuraatheid van FCD is van kritieke belang om verkeersinligting te evalueer vir gebruik van belanghebbendes wat betrokke is by vervoerbeplanning. Hierdie studie is uitgevoer om die akkuraatheid van spoed-data, wat deur kommersiële FCD in landelike gebiede gerapporteer word, te meet. Die studie het maatstafspoeddata en TomTom-historiese snelhede op 12 landelike roetes en 7 stedelike roetes in die Wes-Kaap in Suid-Afrika vergelyk. Die maatstafdata is verskaf deur die WesKaapse Departement van Vervoer, ingesamel deur induksie-lusse. Data wat ingesamel is, is ontleed vir weeksdae (Maandag tot Vrydag) in Februarie 2019 en saamgevoeg oor 1-uur tydsintervalle. Dit is gedoen om 'n ware voorstelling van verkeerspatrone te gee en verkeerspatrone voor die COVID-19- pandemie te ontleed. Om akkuraatheid te evalueer, is toestelpenetrasiekoers, “signed error bias” (SEB), die gemiddelde absolute spoed verandering (AASE), en persentasie vout bepaal. Die minimum toestelpenetrasiekoers, wat vir akkuraatheidsmetings oorweeg is, was 4% en die maksimum toelaatbare foute vir SEB, AASE en persentasie fout was ±7.5 km/h, ±10 km/h en ±10 % onderskeidelik. Die SEB- en AASEakkuraatheidsmaatreëls het getoon dat FCD-gerapporteerde snelhede laer in stedelike gebiede was, terwyl sommige roetes in die landelike gebiede hoër FCD-spoede as die maatstafdata aangeteken het. Toestelpenetrasiekoerse vir beide stedelike en landelike gebiede het aangedui dat 'n hoë toestelpenetrasiekoers nie direk eweredig is aan 'n hoë vlak van spoeddata-akkuraatheid nie. Hierdie studie is relevant omdat evaluering van kommersiële FCD-akkuraatheid nie voorheen in landelike gebiede ondersoek is nie. Hierdie insiggewende navorsing kan help met die verbetering van die vervoerbedryf regoor die wêreld en veral in ontwikkelende lande.en_ZA
dc.description.versionMastersen_ZA
dc.format.extentx, 77 pages : illustrations.en_ZA
dc.identifier.urihttp://hdl.handle.net/10019.1/127303en_ZA
dc.language.isoen_ZAen_ZA
dc.language.isoen_ZAen_ZA
dc.publisherStellenbosch : Stellenbosch Universityen_ZA
dc.rights.holderStellenbosch Universityen_ZA
dc.subject.lcshIntelligent transportation systems -- Data processingen_ZA
dc.subject.lcshTraffic patternsen_ZA
dc.subject.lcshTraffic flowen_ZA
dc.titleA study to measure the accuracy of speed data reported by floating car data in rural areasen_ZA
dc.typeThesisen_ZA
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
kiautha_study_2023.pdf
Size:
3.31 MB
Format:
Adobe Portable Document Format
Description:
Download
Collections
Conference Proceedings (Civil Engineering)