Artificial Intelligence for Smart Transportation

Additionally, new on-demand modalities including ride-share, bike-share, and e-scooters have been introduced in recent years and transformed the transportation landscape in urban environments. A wellfunctioning transit system fosters the growth and expansion of businesses, distributes social and economic benefits, and links the capabilities of community members, thereby enhancing what they can accomplish as a society [6, 11, 15]. However, the explosion in transportation options and the complicated relationship between public and private offerings present myriad new challenges in the design and operation of these systems. There are also complex, and often competing, operational objectives that complicate the implementation of efficient services. Since affordable public transit services are the backbones of many communities, solving these problems and understanding state-of-the-art methods for AI-driven smart transportation has the potential to strengthen urban communities, address the climate challenge, and foster equitable growth. Fundamentally, the design of a well-functioning transit system requires solving complex combinatorial optimization problems related to planning and real-time operations. These problems span many well studied fields, from classical line planning to offline and online vehicle routing problems (VRPs). While there are many ways to assess the performance of smart transportation systems, we largely focus on evaluating these systems in the context of optimizing utilization (i.e.