CSR: Small: Data Services for Reliable Crowdsensing in Urban Spaces

该项目开发了可靠的人群感知理论和工具，旨在大幅降低将城市空间转变为未来智慧城市所需的基础设施投资成本。在本项目中，群体感知是指利用城市中最聪明的资产--人--来分享他们环境中的数据，从而减少对昂贵的物理传感器的需求。作为一个例子和活生生的测试平台，这项工作建立了一个智能交通应用程序，提供了一种新的基于车辆众测的导航服务，称为GreenGPS。这项服务的新奇之处在于为司机找到最省油的路线，而不是最短或最快的路线，根据初步测试，节省了10-20%的能源成本。 据估计，今天，54%的世界人口生活在城市。到2050年，这一比例将增加到66%。城市居民是该项目的主要利益攸关方和受益者。他们是智能交通应用中的驾驶员，灾难响应场景中的幸存者，以及节能系统中的消费者。他们拥有越来越多的传感平台，如相机、GPS设备、互联网连接的汽车、智能手机和活动监测可穿戴设备。通过构建工具，使智慧城市应用广泛依赖于选择参与的人类利益相关者，该项目可以利用现有的传感器池，促进交通流量、能源消耗或灾后管理等领域的城市规模决策。一个关键的分析贡献在于发展的数学基础上实现信息的可靠性保证的基础上，一般不可靠的数据，人类来源可能提供。该项目还包括为不同层次的本科生和研究生教学创建关于下一代众测系统的教育模块。该项目将研究和教育相结合，旨在促进城市空间向智慧城市的发展，同时教育未来的决策者和劳动力了解转型的理论和工具。

This project develops theory and tools for reliable crowdsensing, aiming to significantly reduce infrastructure investment costs needed to turn urban spaces into future smart cities. Crowdsensing, for purposes of this project, refers to the utilization of the smartest asset in the city - the people - in sharing data on their environment, thereby reducing the need for expensive physical sensors. As an example and living testbed, the work builds a smart transportation application, offering a novel vehicular crowdsensing-based navigation service, called GreenGPS. The novelty of this service is in finding the most fuel-economic routes for drivers, as opposed to the shortest or fastest, saving 10-20% in energy costs according to initial tests. It is estimated that, today, 54% of the world population lives in cities. This percentage will increase to 66% by 2050. City dwellers are the primary stakeholders and beneficiaries of this project. They are the drivers in smart transportation applications, the survivors in disaster response scenarios, and the consumers in energy saving systems. They are in possession of increasingly many sensing platforms, such as cameras, GPS devices, Internet-connected cars, smartphones, and activity monitoring wearable devices. By building the tools that broadly enable smart city applications to rely on opt-in human stakeholders, the project can leverage the pool of their existing sensors to facilitate informed city-scale decision-making in areas like traffic flow, energy consumption, or post-disaster management. A key analytic contribution lies in developing the mathematical foundations for attainment of information reliability guarantees on top of the generally unreliable data that human sources may provide. The project also includes creating educational modules for different levels of undergraduate and graduate instruction on the topic of next generation crowdsensing systems. Combining research and education, the project aims to facilitate evolution of urban spaces into smart cities, while educating future decision-makers and workforce on theory and tools underlying the transition.