Clowder Helps SMU Researchers Identify Infrastructure Deserts in Dallas April 1, 2022 In the News EngineeringSoftware and Applications Share this page: Twitter Facebook LinkedIn Email SMU Professor Barbara Minsker and SMU graduate student Zheng Li visit one of the Dallas infrastructure deserts they identified. Their research was turned into a series of interactive reports. Photo credit: SMU By Sophie Anh Bui SMU researchers used drones, machine learning and Clowder, an open-sourced, customizable and scalable web-based data management framework developed at NCSA, to identify 62 Dallas neighborhoods that are highly deficient in infrastructures such as sidewalks, internet, parks, grocery stores, hospital access and more. These deficiencies, also called “infrastructure deserts,” adversely affect the quality of life for people living in these neighborhoods. This study sheds light on the racial inequities in housing and health, specifically in the southern part of Dallas, whose residents are mostly low-income Black and Hispanic families. It emphasizes the need to increase funding for smaller-scale infrastructure projects that seek to make life-enhancing improvements for these neighborhoods and communities. In the SMU study on Dallas’ infrastructure deserts, neighborhoods seen in dark red are “highly deficient.” Circles denote the predominant race-ethnicity in the area. Blue for white, green for Black, reddish for Hispanic, purple for no predominance. Image credit: SMU SMU Civil and Environmental Engineering Professor Barbara Minsker supports this study using a 5-year National Science Foundation grant that seeks to enhance Clowder’s core systems to benefit large user groups. This framework is a sustainable software resource that supports convergent research data needs and preservation across multiple domains – which this study requires. Clowder allows Minsker’s team to easily store complex datasets and share them with the research community. From a data-management perspective, Minsker’s use case on infrastructure deserts is very compelling because of the breadth of data and the algorithms used for data fusion. In the past, we have worked with her team to develop new machine-learning models to analyze a subset of the data and create automatic data pipelines. In the future, we hope to help her team simplify the process of aggregating, distributing and fusing data for new cities by leveraging Clowder and its geospatial extensions.Luigi Marini, NCSA Lead Research Programmer, Clowder Lead Architect. Minsker leads a team of 20+ civil engineering researchers alongside SMU Ph.D. student Zheng Li. Together they’ve created a computer framework to assess 12 types of neighborhood infrastructures and allows the team to evaluate, compare and rate these areas by deficiency, income and ethnicity to identify vulnerable areas. Although this study primarily focuses on Dallas, the team leans on Clowder to allow other cities to access the algorithms and analysis tools to determine their own infrastructure deserts and deficiencies. Read more about this study in a feature story by Dallas Innovators and a recent Clowder newsletter.