The growing complexity of the urban travel pattern and its related traffic congestion, along with the extensive usage of mobile phones, invigorated On-Demand Mobility Services (ODMS) and opened the door to the emergence of Transportation Network Companies (TNC). By adopting the shared economy paradigm, TNCs enable private car owners to provide transportation services to passengers by providing user-friendly mobile phone applications that efficiently match passengers to service providers. Considering the high level of flexibility, convenience, and reliability of ODMS, compared to those offered by traditional public transportation systems, many metropolitan areas in the United States and abroad have reported rapid growth of such services. This dissertation presents a modeling framework to study the operation of on-demand mobility services (ODMS) in urban areas. The framework can analyze the operation of ODMS while representing emerging services such as ridesharing and transfer. The problem is formulated as a mixed-integer program and an efficient decomposition-based methodology is developed for its solution. This solution methodology aims at solving the offline version of the problem, in which the passengers’ demand is assumed to be known ii for the entire planning horizon. The presented approach adopts a modified column generation algorithm, which integrates iterative decomposition and network augmentation techniques to analyze networks with moderate size. Besides, a novel methodology for integrated ride-matching and vehicle routing for dynamic (online) ODMS with ridesharing and transfer options is developed to solve the problem in real-time. The methodology adopts a hybrid heuristic approach, which enables solving large problem instances in near real-time, where the passengers’ demand is not known a priori. The heuristic allows to (1) promptly respond to individual ride requests and (2) periodically re-evaluate the generated solutions and recommend modifications to enhance the overall solution quality by increasing the number of served passengers and total profit of the system. The outcomes of experiments considering hypothetical and real-world networks are presented. The results show that the modified column generation approach provides a good quality solution in less computation time than the CPLEX solver. Additionally, the heuristic approach can provide an efficient solution for large networks while satisfying the real-time execution requirements. Additionally, investigation of the results of the experiments shows that increasing the number of passengers willing to rideshare and/or transfer increases the general performance of ODMS by increasing the number of served passengers and associated revenue and reducing the number of needed vehicles.

Degree Date

Spring 5-14-2022

Document Type


Degree Name



Civil and Environmental Engineering


Khaled Abdelghany

Subject Area

Civil Engineering

Number of Pages




Creative Commons License

Creative Commons Attribution-Noncommercial 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License