LINKING SUSTAINABILITY, RESILIENCE, AND LIVABILITY WITH SMART CITY DEVELOPMENT: BUILDING A NOVEL HYBRID DECISION SUPPORT MODEL FOR COMPOSITE PERFORMANCE ASSESSMENT

Abstract

Smartening development in cities have reinvented hopes to melt down predicaments in early 2000s'. At the embryonic stage, it is vital for cities of today to gain a more consistent understanding on how resilience, livability, and sustainability can be co-created into smart city planning models under a unified umbrella. In that respect, this dissertation attempts to understand smart city development through the lens of sustainability, urban resilience, and livability by proposing a novel hybrid decision support performance assessment model, as cities evolve to achieve the descriptive goal of "Futuristic cities". State-of-the art contribution of this dissertation brings in-house novelty in terms of the subject handled and the approach used to solve the problem. The hybrid decision support model brings in; systems thinking, non-parametric optimization-based envelopment analysis, explainable machine learning based assessments and multi-criteria based combinatorial evaluations all under a unified frame at various levels of measurement. Systems thinking aids in understanding the complexities from a non-fragmented system-of-system perspective. A double-frontier slacks-based measure data envelopment analysis model, a true input-output desirability inclusion model under extended strong disposability assumptions is proposed to evaluate the sustainability performance of smart cities. A relative multivariate metric distance-based approach is proposed to weight the indicators across various dimensions of resilience and livability combining machine learning techniques. Then, the extended version of the Evaluation Based on Distance from Average Solution (EDAS) method combined under a spherical fuzzy (SF) environment with the Analytic Hierarchy Process (AHP) is used to select the best performing smart city and rank them based on the triple criteria of futuristic smart cities (sustainability + resilience + livability). This marks the development of the aspiring "Futuristic Smart City" (FSC) composite index. Data from 35 European smart cities ranked in the top 50 global best smart cities list is taken to empirically evaluate the sustainability, resilience, livability, and their unified performance. The results of the non-parametric optimization-based envelopment analysis revealed significant difference in the productivity progress values from the optimistic and pessimistic viewpoint, thus exemplifying the significance for the proposed aggregate productivity progress measurement model. The results of the machine learning based assessment revealed Gradient Boosting Machine (GBM) as the best classification and predictive model for the resilience, liveability, and aggregate performance assessment. The composite index proposed through the SF-AHP & extended EDAS method revealed London as the top ranked smart city that co-create sustainability, resilience, and livability holistically into their development model. Dusseldorf, Zurich, Munich, Oslo, Dublin, Amsterdam, Hamburg, Rome, Moscow, and Stockholm were no exemption to addressing the triple criteria. The proposed hybrid model augments planned decision making and policy constitution from a strategic level for urban planners and smart city development authorities to support the meta goal of futuristic cities in tech-driven intelligent living units. Tailored towards data-driven and intelligent system approaches, the findings of this dissertation finds applicability not only in the proposed regions as case studies, but in a global scale for any aspiring cities that intent for transition towards futuristic cities.