Abstract
This doctoral study presents a novel, AI-powered framework designed to facilitate the sustainable transformation of low-income cities into climate-resilient, smart urban environments. Central to this research is the imperative to address the mounting challenges faced by resource-constrained urban areas in achieving sustainability goals, amidst increasing climate vulnerabilities, governance limitations, infrastructural deficiencies, and economic constraints. Unlike prevailing smart city models—predominantly constructed for high-income, technologically advanced contexts—this research advances a new paradigm that democratises AI technologies for application in low-income urban settings. The framework is purposefully aligned with the United Nations Sustainable Development Goals (UNSDGs), particularly SDG 11 (Sustainable Cities and Communities), SDG 13 (Climate Action), and SDG 9 (Industry, Innovation and Infrastructure), ensuring its theoretical grounding in globally recognised sustainability standards.
The core contribution of this study lies in the development, validation, and simulation of an AI-centric framework that is both modular and adaptable to the varied realities of low-income cities. The research highlights specific AI applications capable of transforming urban governance, infrastructure, and service delivery systems, with empirical focus areas including traffic management, energy optimisation, predictive infrastructure maintenance, environmental monitoring, and water resource management. Through the integration of real-time sensor data, AI-driven traffic management systems were found to significantly reduce congestion and vehicular emissions in pilot simulations. Similarly, AI-enabled energy management applications, based on predictive analytics and machine learning, demonstrated measurable gains in energy efficiency across building systems—especially in scenarios where traditional energy monitoring mechanisms were absent or unreliable.
Another notable finding relates to the implementation of predictive maintenance algorithms, which enable city authorities to detect and address infrastructure weaknesses before critical failures occur. This capability is particularly valuable for cities with chronic underinvestment in infrastructure and limited operational budgets. The framework also includes AI-driven environmental monitoring tools capable of conducting continuous real-time assessment of air and water quality, thereby providing early warning systems for pollution and enhancing the city's capacity for public health interventions. Across these application areas, the study identified and applied 165 KPIs, systematically extracted from literature, case studies, and scenario modelling, to evaluate the efficacy and scalability of AI tools within climate action frameworks.
Methodologically, this research employs a rigorously designed mixed-methods approach, integrating elements of pragmatism and critical realism to ensure theoretical depth and empirical applicability. The study triangulates findings from a systematic literature review, comparative case study analysis, simulation modelling, and quantitative correlation and factor analysis. Four cities—Sana’a, Baghdad, Dubai, and London—were selected as comparative case studies to capture a spectrum of urban development contexts, ranging from highly resource-constrained environments to advanced digital economies. Case analysis revealed that cities such as Sana’a and Baghdad are impeded by institutional fragmentation, political instability, and infrastructural decay, which constrain the adoption of AI technologies. In contrast, Dubai and London provided models of best practice, particularly in the domains of AI-enabled mobility, decentralised energy systems, and integrated urban planning.
Simulation models were utilised to validate the framework under different climate change and governance scenarios. These simulations enabled the identification of high-priority variables—such as traffic flow reduction, CO₂ emission levels, energy usage, and predictive maintenance frequency—thus offering quantifiable projections of AI’s effectiveness. The outputs indicate that phased, KPI-guided deployment of AI systems can significantly enhance urban resilience and operational efficiency, even in contexts where digital infrastructure and data availability are limited. Moreover, the study’s use of open-source simulation environments and replicable data structures ensures that the framework is both scalable and transferable across diverse urban contexts.
This research contributes to theory by advancing the understanding of AI's role in socio-technical urban transformations, particularly in low-income cities. It challenges the epistemological bias inherent in much of the smart city literature, which tends to exclude informal urbanism and decentralised governance models. Theoretically, the study aligns AI-driven urban innovation with systemic change models, ethical AI governance, and climate adaptation theory. It also integrates frameworks such as the Circular Economy Model (CEM), the Inclusive Smart Framework (ISF), and the Triple Bottom Line (TBL), ensuring that technological development is embedded within broader social, economic, and environmental considerations.
Practically, the research offers actionable recommendations for urban policymakers, planners, and technology developers. It proposes a phased AI adoption roadmap that begins with low-cost, high-impact interventions and progressively integrates more complex. The study also highlights the need for participatory governance structures, transparent algorithmic oversight, and targeted capacity-building initiatives to ensure that AI technologies are deployed ethically and equitably. By identifying both enabling factors and systemic barriers, the research provides a roadmap for overcoming institutional inertia and financial constraints in the AI implementation process.
In conclusion, this study makes a substantial theoretical, methodological, and applied contribution to the growing body of knowledge on AI and urban sustainability. It establishes a new research trajectory that centres the needs of low-income cities in global smart city discourse and demonstrates that artificial intelligence, when contextualised and deployed responsibly, can serve as a powerful catalyst for climate-resilient, inclusive, and data-driven urban development. The framework proposed herein offers both a blueprint and a policy toolkit for cities seeking to align technological innovation with sustainable development imperatives, ensuring that no city is left behind in the global transition toward smart, resilient urban futures.