UHDNetCity: User-centered Heterogeneous Data Fusion for Multi-networked City Mobility

随着世界上越来越多的城市遭受交通拥堵、污染和能源开发的困扰，随着人口增长和城市化进程的继续，城市交通仍然是城市面临的最严峻挑战之一。到目前为止，最常用的城市机动性表征方法主要集中在车辆的空间和时间位置。然而，城市流动性是城市生活的多维特征，在时空框架内经历了围绕着人们及其需求的相互关联的基础设施和信息网络的错综复杂的层层。因此，对移动性的研究应该超越交通系统，以客户为中心，并融入其他物理系统和网络。这一早期概念探索性研究补助金(AGIRE)项目的动机是需要提高对城市流动性的情景意识，并向城市管理者和城市居民分发与城市流动性相关的问题的可靠和及时的信息。通过成功的合作，该项目旨在开发具有可测量指数的城市流动性的新定义，以表征围绕整合交通网络、电力网络和众包数据的公民的城市流动性范式。这个热切的项目预计将有助于该团队与佛罗里达州塔拉哈西市合作，在全球城市团队挑战赛(GCTC)中既定和持续的努力。研究小组已经完成了GCTC的第一阶段，这一迫切的项目将通过开发一种数据驱动的方法来表征城市机动性，从而为第二阶段奠定基础，该方法整合了从塔拉哈西市内的交通网络、电力网络、天气、空气质量和社交媒体收集的数据。这一方法将使塔拉哈西市通过可衡量的机动性好处改善城市服务，朝着成为“智能城市”的努力更近一步，并提高居民的生活质量。这一方法将得到积极的GCTC行动组的支持，其中包括Internet2、EDD Inc.和Stantec公司，以支持塔拉哈西GCTC的努力。UHDNetCity将能够带来可衡量的移动性好处，并从以下方面改善塔拉哈西居民的生活质量：(1)降低车辆和基础设施的能源消耗，(2)减少拥堵、撞车和旅客挫折感，(3)提高安全性和可靠性，(4)提供更精简、更高效和更具成本效益的系统来运营和维护城市服务网络。UHDNetCity框架结合了数据融合、信号处理和机器学习，通过处理不同时空测量数据和网络模型，为实时城市流动性感知提供统一的数学基础。这一数学框架将导致弥合监督和半监督机器学习算法之间的差距，以利用异类城市数据源中的隐藏数据结构来表征城市机动性。UHDNetCity采用用户驱动、以游戏为中心的设计方法，鼓励居民采用塔拉哈西市开发的数字移动应用程序等城市众包仪表盘，并促进他们参与城市交通管理。

As more of the world's cities suffer from congestion, pollution, and energy exploitation, urban mobility remains one of the toughest challenges that cities face as the process of population growth and urbanization continues. So far, the most common approach for urban mobility characterization focuses on vehicle's spatial and temporal positions. However, urban mobility is a multidimensional characteristic of the city life, experienced as tangled layers of interconnected infrastructures and information networks around people and their needs in a spatio-emporal frame. As a result, the study of mobility should go beyond transportation systems, be customer-centered and merged into other physical systems and cyber networks. This Early-concept Grant for Exploratory Research (EAGER) project is motivated by the need to increase the situational awareness in urban mobility and distribute reliable and timely information to city managers and city residents about issues associated with urban mobility. Through successful collaboration, this project aims to develop a new definition of urban mobility with measurable indices to characterize the urban mobility paradigm around citizens integrating transportation networks, electricity networks, and crowdsourced data. This EAGER project is expected to contribute to the team's established and ongoing effort in the Global City Teams Challenge (GCTC) in collaboration with the City of Tallahassee, Florida. The research team has completed the first phase of the GCTC, and this EAGER project will lay the foundation for the second phase by developing a data-driven approach to characterize urban mobility, which integrates collected data from the transportation network, electricity network, weather, air quality and social media within the City of Tallahassee. This approach will put the City of Tallahassee one step closer in their efforts towards being a "smart city" by improving the city services through measurable mobility benefits, and enhance the quality of life for residents. This approach will be supported by the active GCTC action cluster including Internet2, EDD Inc., and StanTec companies to support the Tallahassee GCTC efforts. The UHDNetCity will be able to bring measurable mobility benefits and improve Tallahassee resident's quality of life in terms of (1) lowering energy consumption by vehicles and infrastructure, (2) reducing congestion, crashes and traveler frustration, (3) improving safety and reliability, and (4) providing a more streamlined, efficient and cost-effective system to operate and maintain city service networks. The UHDNetCity framework combines data fusion, signal processing, and machine learning, to provide a unified mathematical foundation for real-time urban mobility sensing by processing heterogeneous spatio-temporal measurement data and network models. This mathematical framework will lead to bridging the gap between supervised, and semi-supervised machine learning algorithms for urban mobility characterization using hidden data structures in the heterogeneous urban data sources. The UHDNetCity employs a user-driven play-centric design approach to encourage resident's adoption of the urban crowdsourcing dashboards such as DigiTally mobile app developed by the City of Tallahassee and promotes their engagement in the urban mobility management.