Impacts of Travel Demand Information Diffusion on Reducing Empty Vehicle Miles Traveled by Ridesourcing Vehicles
Ride-hailing or ride-sourcing services provide alternative mobility options in several cities and have seen exponential growth in market share owing to the convenience they provide. Even though usage of such services may be associated with car-light and car-free lifestyles, there are growing concerns around their impact on urban transportation network operations due to empty miles driven without a passenger (deadheading). This study is motivated by the potential for reducing deadhead mileage by providing information on future trip demand to ride-sourcing drivers (at the destination end of their current trip). Future trip demand information enables the driver to wait in place for the next rider without cruising around and contributing to congestion. A machine learning model is applied using historical data to predict hourly and 10-minute future interval demand for ride-sourcing. Two heuristic algorithms have been developed to i) determine drivers that act on demand information received by waiting in place for the next rider, and ii) match drivers who wait to their next trip by minimizing deadheading distance. Data from ride-sourcing providers in Austin, Texas (RideAustin) and in Chengdu, China (DiDi) are leveraged to apply the methodology. Results show that the process results in 68%-82% and 53%-60% trip-level deadheading miles reduction for RideAustin and DiDi sample operations, respectively. As the maximum tolerable driver's waiting time increases, deadheading savings slightly increase. Even if the maximum percentage of drivers waiting is reduced, significant average deadheading savings by trip can be achieved for that portion.
- Research Organization:
- National Renewable Energy Lab. (NREL), Golden, CO (United States)
- Sponsoring Organization:
- USDOE National Renewable Energy Laboratory (NREL), Laboratory Directed Research and Development (LDRD) Program
- DOE Contract Number:
- AC36-08GO28308
- OSTI ID:
- 1494067
- Report Number(s):
- NREL/PR-5400-73061
- Resource Relation:
- Conference: Presented at the Transportation Research Board (TRB) Annual Meeting 2019, 13-17 January 2019, Washington, D.C.
- Country of Publication:
- United States
- Language:
- English
Similar Records
A Deep Learning Approach for TNC Trip Demand Prediction Considering Spatial-Temporal Features: Preprint
Data-Driven Multi-Step Demand Prediction for Ride-Hailing Services Using Convolutional Neural Network