EAGER: Agile Data Integration to Facilitate Scaling of Air Quality Research

交通运输通过车辆排放对城市地区的空气污染产生重大影响，给行人、车辆乘员和过境旅客带来健康风险。道路附近的空气污染物浓度可能比城市地区的平均空气污染物水平高出几个数量级。此外，根据美国环保署的数据，交通运输产生的温室气体占美国总排放量的26%。波特兰州立大学最近与俄勒冈州波特兰市合作进行的一项研究表明，一种相对简单的技术——修改交通信号的时间——有可能减少城市中车辆的排放。然而，这项初步研究需要更充分地探索，以评估其潜力。该项目将把工作从一个地点扩展到一个完整的运输走廊，即鲍威尔-迪格走廊。此外，该项目将研究如何将数据管理技术应用于空气质量分析。我们探索性研究得出的技术有望为在多个领域推进“智慧城市”方法的努力提供信息。拟议的项目利用独特和时间敏感的资源和机会，设计一种创新和潜在的变革方法，以解决与全球相关的问题-减少与交通有关的空气排放。通过确定一种相对容易实施的方法来减少车辆排放，从而减少温室气体排放，改善空气质量，减少行人接触空气污染物，拟议的工作有可能影响城市居民的生活。该项目直接为波特兰（俄勒冈州）全球城市挑战（GCTC）行动集群做出贡献，该行动集群最近在2016年GCTC博览会上获得了20,000美元的领导奖。该项目还与波特兰市的波特兰无处不在的移动（UB Mobile PDX）计划保持一致，该计划是美国交通部智能城市挑战赛的七个决赛入围者之一。该项目将调查和开发网络物理系统技术，特别是数据管理技术，以解决在空气质量和运输数据的数据清洗和数据集成中观察到的系统问题。在实践中，数据集成和清理通常仍然以特别的方式自动化；现有的系统数据集成和清理技术不能有效地支持这些过程的扩展。该项目提出了一个我们称之为敏捷集成的概念，旨在解决与环境传感数据快速增长、城市变化速度加快以及数据密集型决策压力增加相关的复杂动态。还将开发半自动数据清理和处理技术，以便更好地捕捉数据集成中的人工决策和判断。研究结果将在用于密集传感器网络的数据集成的网络物理系统原型中实现。从长远来看，拟议的工作有可能通过验证通过改变信号定时来减少车辆排放的潜在方法，从而影响日常市民的生活。城市车辆排放影响温室气体排放和城市空气质量。由于少数民族和社会经济地位低的人口不成比例地居住在靠近主要道路的地方，本项目的潜在影响直接影响到那些往往得不到充分服务的人口。此外，上文所述的可扩展性问题，虽然以本提案的空气质量研究为例，但也出现在交通领域以及医疗保健、教育和环境传感等其他领域。因此，通过该项目开发的技术有望扩展到这些领域。这项工作将提高人们对低成本空气质量传感器的理解。就教育目标而言，该项目将吸引来自PSU？这是一个专门招募美国原住民和俄勒冈州农村居民的大气科学REU。结果将传播到计算机科学、交通和空气质量专业社区，从而影响至少三个研究领域。

Transportation, through vehicle emissions, has a significant impact on air pollution in urban areas - presenting health risks to pedestrians, vehicle occupants and transit users. Air pollutant concentrations near roadways may be up to orders of magnitude higher than average air pollutant levels in urban areas. Further, according to the EPA, transportation accounts for 26% of greenhouse gasses in the United States. Recent research at Portland State University, in collaboration with the City of Portland, Oregon, indicates that a relatively simple technique - modifying the timings of traffic signals - has the potential to reduce vehicle emissions in cities. However, this preliminary research needs to be explored more fully to evaluate its potential. This project would scale the work from a single location to a full transportation corridor, namely the Powell-Division Corridor. In addition, this project will investigate how data management technology can be applied to scale the air quality analysis. The techniques resulting from our exploratory research are expected to inform efforts to advance 'Smart City' approaches in multiple domains. The proposed project capitalizes on unique and time-sensitive resources and opportunities to design an innovative and potentially transformative approach to address a globally relevant problem - reducing traffic-related air emissions. The proposed work has the potential to affect the lives of urban citizens by identifying a relatively easy to implement method for reducing vehicle emissions and thereby reducing greenhouse gas emissions, improving air quality and reducing pedestrian exposure to air pollutants. This project directly contributes to the Portland (Oregon) Global Cities Challenge (GCTC) Action Cluster, recently awarded the $20,000 leadership award at the GCTC 2016 Exposition. The project also aligns with the City of Portland's Ubiquitous Mobility for Portland (UB Mobile PDX) initiative, one of seven finalists in the U.S. Department of Transportation Smart Cities Challenge.This project will investigate and develop Cyber Physical Systems technology, particularly data management technology, to address systematic issues observed in data cleaning and data integration of air quality and transportation data. In practice, data integration and cleaning are still typically automated in an ad-hoc fashion; existing systematic data integration and cleaning technologies do not effectively support scaling of these processes. This project proposes to develop a concept we call Agile Integration, which is designed to address the complex dynamics associated with rapid increases in environmental sensing data, the accelerating pace of change in cities, and mounting pressures on data-intensive decision making. Techniques for semi-automated data cleaning and processing will also be developed to better capture human decisions and judgments that go into data integration. The results will be implemented in a prototype Cyber-Physical System for Data Integration for dense sensor networks. In the long term, the proposed work has the potential to impact the lives of everyday citizens by validating a potential method for reducing vehicle emissions through signal timing changes. Vehicle emissions in urban areas impact greenhouse gas emissions and urban air quality. Since minority and low-socioeconomic status populations disproportionately reside in close proximity to major roadways, the potential impacts of this project directly affect those often underserved populations. Further, the scalability problems described above, while exemplified by the air quality research for this proposal, also appear in the transportation domain and in others such as healthcare, education and environmental sensing. Thus the techniques developed through this project are expected to be extensible to those domains. The work will produce an improved understanding of lower-cost air quality sensors. In terms of educational goals, this project will engage students from PSU?s atmospheric science REU that specifically recruits Native American and rural Oregonians. Results will be disseminated to the computer science, transportation, and air quality professional communities, thus impacting at least three research domains.