A Probabilistic Location Prediction Approach to Optimize Dispatch Processes in the Ride-Hailing Industry

Date
2021-01-05
Authors
Richly, Keven
Schlosser, Rainer
Brauer, Janos
Plattner, Hasso
Contributor
Advisor
Department
Instructor
Depositor
Speaker
Researcher
Consultant
Interviewer
Annotator
Journal Title
Journal ISSN
Volume Title
Publisher
Volume
Number/Issue
Starting Page
1830
Ending Page
Alternative Title
Abstract
For peer-to-peer ride-hailing providers, it is a crucial competitive advantage to cost-efficiently dispatch passenger requests and to communicate accurate waiting times. To determine waiting times and dispatch decisions, transport network companies need precise information about the location of all available drivers. Due to technical limitations and outdated data (e.g., low sample rates, continuous movement of drivers), however, existing systems, which typically use the last observed locations of drivers, regularly suffer from dispatches with critical delays. In this paper, we present an approach to predict probability distributions for drivers' future locations, which are calculated based on patterns observed in past trajectories. We evaluate the applicability and accuracy of the proposed algorithm on a large real-world trajectory dataset of a transportation network company. Our results allow quantifying the risk of critical delays and thus enable risk considerations in improved dispatching strategies.
Description
Keywords
Smart Mobility Ecosystems and Services, location prediction, ride-hailing, trajectory data, transportation network companies
Citation
Extent
11 pages
Format
Geographic Location
Time Period
Related To
Proceedings of the 54th Hawaii International Conference on System Sciences
Table of Contents
Rights
Attribution-NonCommercial-NoDerivatives 4.0 International
Rights Holder
Local Contexts
Email libraryada-l@lists.hawaii.edu if you need this content in ADA-compliant format.