UC Santa Barbara

Spatial analysis, especially spatial optimization, is critical for many applications in urban planning, e.g., facility location, route planning, land use management, among others. Location is the core of methods in spatial analysis, injecting spatial knowledge into the modeling process in either an explicit or implicit way. However, orientation or direction as another important spatial factor in geography is often neglected in literature. Orientation or direction contains rich spatial information, enabling accurate and realistic modeling of spatial objects, relation, and phenomena. Absence of orientation or direction modeling in quantitative analysis can lead to inferior or even erroneous results, impacting efficiency and accessibility. The goal of this dissertation is to explore different ways of incorporating orientation into spatial analysis and modeling, highlighting the important role of orientation from data preprocessing, spatial evaluation to decision making. Specifically, orientation is explicitly considered in various spatial optimization models for point cloud completion, facility location assessment and directional facilities siting. In addition, different solution techniques have been developed to solve these orientation-enhanced models. The contributions of this dissertation are many. First, the challenges of integrating orientation in spatial analysis and modeling are analyzed. Second, orientation is explicitly incorporated into spatial optimization models, providing a new paradigm for spatial analysis. Lastly, new solution techniques are developed and applied to real use cases in support of urban planning.