EAGER/Collaborative Research: Towards Dynamic Social Ride-sharing: An Essential Component in Smart and Connected Communities

点击翻译按钮获取中文摘要

Prior work has demonstrated the potential economic and environmental benefits of ride-sharing; however, ride-sharing has not been adopted by the majority of travelers, since previous efforts aimed to maximize the number of participants but failed to take travelers social attributes into account. One of the major challenges in advocating ride-sharing is how to reduce social barriers. Taking advantage of advances in information technology and social media for building smart communities, and with emerging social attitudes regarding travel behavior, ride-sharing along with its social experience should be considered as a viable and possibly dominant mode of transportation in smart communities. This EArly-concept Grant for Exploratory Research (EAGER) project will address the current social barriers to ride-sharing and will provide more appealing ride matches leading to increased user participation. While a slight compromise in travel time may be tolerated by users in exchange of better social experience, the economy of scale in social ride-sharing would result in better performance of transportation systems in smart cities and communities. The project will support the realization of asocial ride-sharing system that can serve in validating the long-term research goals of deploying Social Transportation. The goal of this research is to develop theoretical foundations for supporting social ride-sharing. Users social attributes and preferences will be jointly determined by two innovative patterns: the spatial-activity pattern and the social-activity pattern. A probabilistic location clustering model will be built to identify abnormal non-commuting travels for the spatial-activity pattern. This project leverages a key word-search approach and topic recognition to estimate travelers social activities and travel purposes for the social-activity pattern. Novel optimization models for estimation of travel pattern from social media and ride-sharing data will be developed, to include using state-of-the-art matrix estimation techniques to infer various aspects of travel pattern, such as regular and special-pattern (anomalous) trips by type, time of day, day-of-week, etc. This step is particularly innovative in that these estimation methods are new to the transportation field, and their successful application would be beneficial to the research community. New theory will be developed for social ride-sharing in maximizing both riders satisfaction and trip efficiency. A living lab will be used in collaboration with local entities to demonstrate the pilot implementation of the research product.