Urban Intelligence: Tools and Models for Smart Cities

Smart city is a concept that is centered around leveraging advancements in technology to solve urban problems. Despite its popularity, many knowledge gaps exist around the concept. The goal of this dissertation is to enhance the understanding of the smart city concept and its implementation practices. Specifically, this dissertation addresses three challenges: (1) the lack of coherent structure in smart city research, (2) the dearth of decision-making tools, and (3) data heterogeneity issues in smart city development. Chapter 1 introduces the dissertation. Chapter 2 tackles the first challenge by analyzing the smart city literature in the context of urban infrastructure. Information related to the studies’ objective(s) are encoded in the form of Knowledge Graphs (KGs) that are later used to identify dominant research themes. The methodology can help researchers infer trends across KGs. In response to the second challenge, Chapter 3 proposes a conceptual model to guide the decision-making process. After identifying the issues associated with the decision-making process in smart city development, the conceptual model of Urban Intelligence that adopts a system-of-systems view is developed and detailed. The model can aid decision-makers/policy makers to assess the effectiveness of solutions to prioritize them. Chapter 4 follows the line of research that uses novel sources of data to identify cities with similar structural properties. Specifically, it analyzes and compares the topological and geometric road network properties of 80 world cities. The analysis yields five clusters of cities; for example, one cluster displays more irregular street patterns whereas another cluster shows more orthogonal street patterns (90 and 180-degree). This approach can help policy makers to study smart city policies implemented in cities that belong to their cluster and determine whether a similar solution could be applied to their city. Finally, Chapter 5 addresses the issue of data heterogeneity. Typically, data from multiple sources are used in parallel to make informed decisions; however, combing those data often poses problems. Chapter 5 studies the challenges encountered in converting Computer Aided Design (CAD) data (a drawing format often used to digitally map infrastructure) to Geographical Information Systems (GIS) data (a more common mapping format) and proposes the C2G conversion framework. The C2G framework can guide anyone with a basic knowledge of CAD and GIS to convert data rapidly with minimal information loss. In summary, while the smart city concept shows promise, many challenges remain. This dissertation identifies several challenges that hamper progress in both smart city research and development, and proposes methodologies to address them, bringing us closer to the smart city.