Identifying and Planning for Group Travellers in On-Demand Mobility Models

Abstract

Understanding group travel is vital for transportation planners and policymakers, especially when modelling emerging on-demand mobility such as ridesharing and shared autonomous vehicles. Existing agent-based simulations of ridesharing services hardly consider group travel, even though these services mainly occur during the weekend and for leisure trips where people are more likely to travel in groups. This is due to the limited availability of group travel data in many travel demand models. This study uses a Swiss synthetic travel demand where car drivers and passengers are modelled separately to identify group travellers. A heuristic approach based on mixed integer linear programming is implemented to create group travellers by matching car drivers and passengers. An agent-based simulation model is set up to simulate ridesharing while considering group travel to reveal the impact on operational policies for ridesharing.