Ridesharing, defined as two or more travelers following similar routes and schedules sharing one car, is usually considered as an effective way to discourage private car ownership. The emergence of ridesharing apps makes it convenient and efficient for strangers to receive real-time matching service of ridesharing. However, while a ridesharing scheme may encourage some travelers to forgo their vehicle purchase plans, it may also induce some others, who originally could not afford the cost of car usage, to purchase cars due to the reduced cost of driving. This thesis mathematically models the long-term impacts of ridesharing on travelers’ vehicle purchase behavior, considering the complex interactions between the attractiveness of ridesharing in terms of matching probabilities a...[ Read more ]

Ridesharing, defined as two or more travelers following similar routes and schedules sharing one car, is usually considered as an effective way to discourage private car ownership. The emergence of ridesharing apps makes it convenient and efficient for strangers to receive real-time matching service of ridesharing. However, while a ridesharing scheme may encourage some travelers to forgo their vehicle purchase plans, it may also induce some others, who originally could not afford the cost of car usage, to purchase cars due to the reduced cost of driving. This thesis mathematically models the long-term impacts of ridesharing on travelers’ vehicle purchase behavior, considering the complex interactions between the attractiveness of ridesharing in terms of matching probabilities and travelers’ mode choices under different car ownership status. Under any cost sharing strategy, a group of inequalities established the equilibrium state. Based on the model, we show by numerical experiments that a ridesharing program may either slow down or accelerate the increase of car ownership, depending on the choice of driver-rider cost sharing strategies. When the variance of ridesharing matching probabilities due to the variance of car ownership is ignorable, we establish the properties of cost sharing strategies that can slow down the increase of car ownership.

As the emergence of autonomous vehicles (AVs) on road, in the transportation network there will be a stage with mixed ridesharing program with AVs and traditional vehicles (TVs). We investigate the morning commute problem with three travel modes: driving alone, traditional vehicles ridesharing (TV-RS) and autonomous vehicles ridesharing (AV-RS). Assuming the compensation scheme is a linear function of shared distance and travel time, two sides of commuters in the TV-RS program, i.e., the TV-RS driver and TV-RS rider, can reach an internal equilibrium according to an appropriate compensation scheme. While under three different provision regimes of AVs, i.e., private owner provision, monopoly provision, and public provision, the corresponding compensation schemes are proposed to ensure a positive ridership in AV-RS program. With well-designed compensation schemes, the equilibrium among the three travel modes will be governed by the bottleneck model to minimize all commuters’ travel cost regardless their travel mode. Under public provision of AVs, we examined the system optimum (SO) condition and investigated the toll strategies to eliminate the queuing delay.