On-demand ridepooling systems usually need to decide which requests to serve, when the number of vehicles is not enough to transport them all with waiting times that are acceptable by the users. When doing so, they tend to provide uneven service rates, concentrating rejections in some zones within the operation area. In this paper, we propose two techniques that modify the objective function governing the assignment of users to vehicles, to prioritize requests originated at zones that present a relatively large rejection rate. The goal is to diminish the Gini Index of the rejections' rate, which is a well established way to measure inequality in economics. We test these techniques over an artificial small network and a real-life case in Manhattan, and we show that they are able to reduce the Gini Index of the rejection rates. Moreover, the overall rejection rate can be simultaneously reduced, thanks to utilizing the vehicles more efficiently.
Towards a geographically even level of service in on-demand ridepooling
2021-09-19
2082198 byte
Conference paper
Electronic Resource
English
Anticipatory routing methods for an on-demand ridepooling mobility system
Online Contents | 2021
|Revealed demand patterns of people with disabilities in on-demand ridepooling
DOAJ | 2025
|Impact of Service Design on Urban Ridepooling Systems
IEEE | 2020
|DataCite | 2021
|Analyzing Urban Residents’ Appraisal of Ridepooling Service Attributes with Conjoint Analysis
German Aerospace Center (DLR) | 2018
|