A recent report funded by the Pacific Southwest Region (PSR) University Transportation Center has expanded the capacity of transportation planners to model the complexity of modern commute choices. For workers, ride-hailing or ride-sourcing companies have simplified commutes by expanding options beyond a private vehicle, carpooling, or public transit. Workers can now commute via ride-hail in the morning and carpool home (or vice versa). This innovation for workers has proven to be a headache for transportation planners because conventional transportation planning models cannot couple, at the individual level, decisions for the morning and evening commute. PSR researchers Jong-Shi Pang, Maged Dessouky, Wei Gu (USC) and Michael Zhang (UC Davis), address that shortcoming by developing a model that considers the joint travel decisions and interactions between solo driving, rideshare (e.g., carpooling) and ride-sourcing in a coupled morning-evening commute modeling framework. Their approach was to develop a general equilibrium model that (i) provides simultaneously the results of traffic flows and travelers’ mode choices; (ii) quantifies the possible mode switches across various transportation services between the morning and evening commutes; (iii) allows for passengers from different Origin-Destination (OD) pairs to share a ride together; and (iv) models the coupling interaction effects between morning and evening commutes. 

The research team validated the model using the well-studied network. Their numerical examples show that considering morning and evening commutes separately tends to overestimate the travel cost, number of drivers (by 18.2 percent) and total Vehicle Hours Traveled (by 8.2 percent) in the network. Their model, which allows individual choices of driving, carpooling, or ride-sharing for both the morning and evening commute, captures the mode switches and interactions between the morning and evening. This leads to fewer drivers and less Vehicle Hours Travelled (VHT) in the system. The implication is that new technologies have the ability to create more flexible choices. For example, persons can be a carpool driver in the morning but drive home solo in the evening, while their carpool partner now can return home via ride-hailing. As other flexibly shared modes – bike-sharing, docked or dockless vehicle sharing, and flexible route paratransit – become more common, these behaviors will also become more common. Their work helps point the way for travel behavior models to catch up, reflecting the complexity of commuting choices at a microscopic (individual) level.

For more information, see the Pacific Southwest Region University Transportation Center report, “Modeling multi-modal mobility in a coupled morning-evening commute framework that considers deadheading and flexible pooling,” available here.