Adequate public spaces and urban green areas are key criteria for urban development and infrastructure implementation in healthy cities. Latterly, there have been an increasing number of research methods using artificial intelligence (AI) to monitor, quantify, and control the state of these spaces with an aim toward pioneering research in urban studies. However, in informal areas, open-data access tends to lack adequate and updated information, making it difficult to use AI methods. Hence, we propose a methodology for restricted open data collection and preparation for future use in machine learning or spatial data science models for similar areas. To that extent, we examine two peripheral and low-income neighborhoods in Quito, Ecuador—La Bota and Toctiuco—to analyze their public spaces, urban green areas, points of interest, and road networks, and how they address healthy cities criteria. We develop an original methodological approach that combines an index of proximity, accessibility, quantity, and quality for these spaces with geospatial and network analysis techniques. Results indicate that the connectivity and structure of these spaces are centralized and nodal, representing exclusion and segregation. This work provides insights into potential healthy spaces and information to urban planners and policymakers in decision-making for healthy urban infrastructure.