Skip to main navigation Skip to search Skip to main content

A USAGE-BASED INSURANCE SYSTEM ACCORDING TO THE CELLULAR-BASED TRAVEL EXPOSURE IN A CRASH LIKELIHOOD MAP

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Usage-based insurance (UBI) taking driving distance and behaviour into account raise a lot of attention in recent years. Unlike most UBI systems, this study proposes a novel exposure-based insurance system comprising a crash likelihood map and travel exposure models. The former, the crash likelihood map, is estimated by a joint crash frequency and severity model based on historical accident spot data and is used to determine the degree of risk in traveling a certain roadway by using private vehicles. The latter, the travel exposure, is then estimated by the Hidden Markov model for travel trajectory and machine learning models for mode classification based on user’s cellular data. To train and validate the proposed cellular-based exposure model, the cellular and travel diary data of 60 volunteers in a whole week were collected. The estimation results for both crash likelihood map and travel exposure models are satisfactory. Additionally, the driving risk of the 60 volunteers are then evaluated by combining the estimated risk map and travel exposure models to show the applicability of the proposed UBI system.

Original languageEnglish
Title of host publicationProceedings of the 27th International Conference of Hong Kong Society for Transportation Studies, HKSTS 2023
Subtitle of host publicationTransport and Equity
EditorsMei-Po Kwan, Sylvia Y. He, Y.H. Kuo
PublisherHong Kong Society for Transportation Studies Limited
Pages576-583
Number of pages8
ISBN (Electronic)9789881581518
StatePublished - 2023
Event27th International Conference of Hong Kong Society for Transportation Studies: Transport and Equity, HKSTS 2023 - Hong Kong, Hong Kong
Duration: 11 Dec 202312 Dec 2023

Publication series

NameProceedings of the 27th International Conference of Hong Kong Society for Transportation Studies, HKSTS 2023: Transport and Equity

Conference

Conference27th International Conference of Hong Kong Society for Transportation Studies: Transport and Equity, HKSTS 2023
Country/TerritoryHong Kong
CityHong Kong
Period11/12/2312/12/23

Keywords

  • Cellular data
  • Crash frequency and severity
  • Map matching
  • Mode classification
  • Usage-based insurance

Fingerprint

Dive into the research topics of 'A USAGE-BASED INSURANCE SYSTEM ACCORDING TO THE CELLULAR-BASED TRAVEL EXPOSURE IN A CRASH LIKELIHOOD MAP'. Together they form a unique fingerprint.

Cite this