Geri Dön

Akıllı sürdürülebilir şehir değerlendirmesine yönelik bir model önerisi

A model proposal for smart sustainable city assessment

PDF İndir

  1. Tez No: 1001056
  2. Yazar: FATMA SENA KARAL
  3. Danışmanlar: DOÇ. DR. AYBERK SOYER
  4. Tez Türü: Doktora
  5. Konular: Endüstri ve Endüstri Mühendisliği, Industrial and Industrial Engineering
  6. Anahtar Kelimeler: Belirtilmemiş.
  7. Yıl: 2026
  8. Dil: Türkçe
  9. Üniversite: İstanbul Teknik Üniversitesi
  10. Enstitü: Lisansüstü Eğitim Enstitüsü
  11. Ana Bilim Dalı: Endüstri Mühendisliği Ana Bilim Dalı
  12. Bilim Dalı: Endüstri Mühendisliği Bilim Dalı
  13. Sayfa Sayısı: Belirtilmemiş.

Özet

Artan kentleşme, şehirlerin yalnızca yaşam alanı olarak değil aynı zamanda sürdürülebilir kalkınmanın merkezi bir aktörü haline gelmesine neden olmuştur. Bu bağlamda, şehirlerin hem akıllı hem de sürdürülebilir olması, günümüz planlama ve yönetişim anlayışının temel hedeflerinden biridir. Ancak bu iki kavramın birlikte ele alındığı değerlendirme çerçeveleri hâlen sınırlıdır. Bu tez, literatürdeki bu boşluğu doldurmak amacıyla, akıllı sürdürülebilir şehir performansını ölçen bütünleşik ve çok aşamalı bir değerlendirme yaklaşımı geliştirmektedir. Çalışma iki temel aşamadan oluşmaktadır. İlk aşamada, Sistematik Literatür Taraması yöntemi kullanılarak akıllı, sürdürülebilir ve akıllı sürdürülebilir şehir kavramlarına dair güncel literatür incelenmiştir. Web of Science, Scopus, ScienceDirect ve Google Scholar veri tabanlarından elde edilen 90 çalışma analiz edilerek şehir değerlendirme süreçlerinde kullanılan performans göstergeleri, boyutlar ve değerlendirme kriterleri çıkarılmıştır. Bu analiz sonucunda ekonomik, çevresel ve sosyal temalar altında sınıflandırılmış 49 gösterge belirlenmiştir. İkinci aşamada, önerilen dört adımlı yaklaşım aracılığıyla şehirlerin akıllı sürdürülebilir şehir performans düzeyleri değerlendirilmiştir. Yaklaşımın ilk adımında, gösterge seçim süreci uzman görüşleri ile yürütülmüş ve 20 gösterge belirlenmiştir. İkinci adımda, göstergelerin göreli önem değerlerini belirlemek amacıyla Entropi ve MEREC yöntemleri birlikte kullanılmış ve çarpımsal normalizasyon yöntemi ile bu iki yöntemden elde edilen ağırlıklar birleştirilmiştir. Üçüncü adımda, performans göstergelerine ilişkin derlenen veriler aracılığıyla her göstergeye ait dilsel skorlar belirlenmiş ve söz konusu skorlar 7 seviyeli bulanık ölçekle ifade edilmiştir. Elde edilen bu ifadeler, Birikimli Kanı Derecesi (CBD) yaklaşımı ve Bonferroni-Sıralı Ağırlıklandırılmış Ortalama (Bon-OWA) operatörü yardımıyla birleştirilerek şehirlerin genel performans düzeyleri hesaplanmıştır. Daha sonra geliştirilen yaklaşım 15 Avrupa şehrine uygulanmış ve elde edilen skorlar kullanılarak ilgili şehirler akıllı sürdürülebilir şehir performans skorlarına göre sıralanmıştır. Paris en yüksek performansı gösteren şehir olurken, Madrid en düşük puanı almıştır. Geliştirilen gösterge paneli aracılığıyla, her şehir için hangi göstergelerin güçlü ya da zayıf olduğu grafiksel olarak sunulmuş ve şehirlerin öncelikli müdahale gerektiren alanları belirlenmiştir. Ayrıca, gösterge ağırlıkları ve dilsel ölçeklerin varyasyonlarına bağlı olarak duyarlılık analizleri gerçekleştirilmiş ve geliştirilen yaklaşımın geçerliliği test edilmiştir. Son olarak, elde edilen sıralamalar, literatürde sıkça kullanılan TOPSIS ve SAW gibi çok kriterli karar verme yöntemleriyle karşılaştırılmıştır. Buna ek olarak, şehirlerin ASŞP skorları global düzeyde kullanılan akıllı şehir ve sürdürülebilir şehir endeksleriyle de karşılaştırılmıştır. Her iki karşılaştırma sonucunda, önerilen yaklaşımın elde ettiği sıralamalar ile diğer yöntem ve endeksler arasındaki benzerlikler ve farklılıklar analiz edilmiştir. Geliştirilen yaklaşım, çok boyutlu ve belirsizlik içeren şehir değerlendirme süreçlerine entegre edilebilecek esnek ve yenilikçi bir metodoloji sunmakta; sürdürülebilir kalkınma hedefleri doğrultusunda şehirlerin mevcut durumlarını kapsamlı biçimde analiz etmeye olanak tanımaktadır. Bu yönüyle model, yalnızca kavramsal düzeyde değil, aynı zamanda uygulamaya dönük yönleriyle de öne çıkmakta; yerel yönetimler, politika yapıcılar ve araştırmacılar açısından stratejik planlama, performans kıyaslama ve iyileştirme alanlarının belirlenmesi gibi karar verme süreçlerinde etkin şekilde kullanılabilecek pratik bir araç niteliği taşımaktadır. Böylece, sürdürülebilir şehircilik alanındaki çalışmalara hem teorik derinlik hem de karar vericilere rehberlik edebilecek somut çözüm önerileri sunarak literatüre ve uygulamaya değerli katkılar sağlamaktadır.

Özet (Çeviri)

The rapid acceleration of urbanization in recent decades has significantly altered the role of cities in global development dynamics. No longer viewed merely as population centers, cities today are recognized as key arenas where the success or failure of sustainability agendas will be determined. As urban populations grow and technological advancements reshape daily life, the demand for governance models that can balance innovation with environmental and social responsibility becomes increasingly urgent. In response to these complex pressures, the concept of Smart Sustainable Cities (SSCs) has emerged as a comprehensive paradigm that seeks to integrate digital transformation with sustainability goals. Despite this growing interest, current academic and practical approaches to assessing SSC performance remain fragmented and limited in scope. Most frameworks either prioritize technological metrics or focus solely on sustainability indicators, rarely achieving a meaningful synthesis of both. This thesis addresses this gap by proposing an integrated and operationally flexible model to assess SSC performance across European cities. The study is methodologically structured into two major phases. The first phase involves a Systematic Literature Review (SLR), designed to establish a solid conceptual and empirical foundation for indicator selection. A total of 90 peerreviewed articles were identified from the Web of Science, Scopus, ScienceDirect, and Google Scholar databases using predefined search queries and inclusion criteria. These studies were published within the last 15 years and span a wide array of academic disciplines, from urban planning and environmental engineering to public administration and information systems. The review focused on how smart, sustainable, and smart-sustainable city concepts have been operationalized, what types of indicators have been used in existing evaluations, and what weighting and scoring mechanisms have been applied. All identified indicators were then categorized according to the Triple Bottom Line (TBL) model, which classifies urban performance under three broad themes: economic, environmental, and social. Based on this analysis, 49 indicators were selected, each representing a key aspect of SSC dynamics. In the second phase of the study, a four-stage decision-making approach was developed to quantitatively evaluate SSC performance. The first stage of the approach focused on refining the initial pool of 49 indicators. Expert feedback was collected from a panel of ten professionals with relevant experience in smart city planning, sustainable urban development, and digital governance. These experts were asked to assess the importance of each indicator and its suitability for comparative evaluation. Based on their responses, indicators with low perceived importance or limited relevance were excluded. A 5% importance threshold, informed by both statistical distribution analysis and Pareto efficiency principles, was applied. As a result, 35 indicators were initially retained. Subsequently, the data availability of these 35 indicators was examined across 15 European cities. To address cases with partial data, the Multiple Imputation by Chained Equations (MICE) method was applied. This statistical technique enabled the estimation of missing values using iterative regression modeling, thereby allowing for a more complete dataset. After applying MICE and excluding indicators with irreparable data gaps, 20 indicators were finalized for use in the subsequent analysis. These indicators were evenly distributed across the three thematic areas—economic, environmental, and social—ensuring both representativeness and thematic balance. In the second stage of the approach, the weighting of indicators was conducted through a hybrid approach combining the Entropy and MEREC methods. The Entropy method captures the variability of each indicator across cities, assigning lower weights to uniformly distributed indicators and higher weights to those that show significant differences. This allows the model to emphasize factors that distinguish urban performance more effectively. In parallel, the MEREC (Method based on the Removal Effects of Criteria) technique evaluates the marginal impact of each indicator on the overall performance when removed from the dataset. Indicators that significantly alter overall scores when omitted are considered more influential and are thus given higher weights. The integration of these two techniques provides a dual perspective—one based on information dispersion, the other based on performance dependency. Rather than averaging the results of Entropy and MEREC, which could obscure their distinct contributions, a multiplicative normalization procedure was employed. In this approach, the individual weights from both methods are multiplied and then normalized across all indicators. This has several methodological advantages. It preserves the sensitivity of each method while penalizing indicators that are weak in both measures. Moreover, it aligns with the theoretical stance that trade-offs between sustainability pillars should be minimized: a poor score in one area (e.g., social equity) should not be fully compensated by a high score in another (e.g., economic growth). By adopting this non-compensatory approach, the model ensures a more conservative and ethically aligned evaluation of urban sustainability. The result of this stage is a set of finalized, objective, and thematically balanced indicator weights. These weights are then carried forward to the subsequent stages of the analysis, where expert-based qualitative data is incorporated and synthesized with the quantitative structure established thus far. In the third stage of the approach, quantitative data were used to assess city-level performance across the 20 retained indicators. For each city–indicator pair, available datasets were collected from official sources, statistical offices, and municipal reports. These raw data were then normalized and translated into performance scores using a 7-point fuzzy linguistic scale ranging from“Very Poor”to“Very Good.”This linguistic transformation was necessary to enable comparability across different indicators measured on diverse scales. Each linguistic term was represented by a Triangular Fuzzy Number (TFN), providing a structured approach to convert numerical values into fuzzy evaluations suitable for aggregation under uncertainty. This step ensured that all indicators—regardless of their original unit or distribution—could be incorporated into a unified multi-criteria evaluation framework. These fuzzy evaluations were then processed using the Cumulative Belief Degree (CBD) method. CBD is particularly well-suited for decision problems involving linguistic uncertainty, expert disagreement, and limited data. Unlike crisp scoring techniques, CBD does not collapse expert input into single point estimates; instead, it represents each performance score as a belief distribution over linguistic categories. This distributional structure allows for preserving uncertainty and enables the aggregation of assessments without discarding minority opinions. For each city and indicator, a belief vector was generated based on the relative agreement across the panel, capturing both the consensus and dispersion in expert views. The final stage of the approach involved aggregating these belief distributions into a composite performance score for each city. To achieve this, the Bonferroni Ordered Weighted Averaging (Bon-OWA) operator was employed. This operator blends the properties of the traditional OWA function with the Bonferroni mean, allowing for non-linear, partially compensatory aggregation. The orness parameter in Bon-OWA was used to adjust the degree of optimism or pessimism in aggregation behavior, reflecting whether the evaluation should reward exceptional strengths or penalize critical weaknesses more heavily. This flexibility was crucial in sustainability assessment, where underperformance in certain areas—such as environmental degradation or social inequality—should carry substantial weight in the final evaluation. After completing the aggregation process, composite performance scores were calculated for each of the 15 European cities included in the study. These scores reflected the integrated performance of each city across all three dimensions: economic vitality, environmental responsibility, and social equity. Paris emerged as the highestranked city, exhibiting consistently strong scores across most indicators, particularly in climate policy, digital infrastructure, and social services. Helsinki, Amsterdam, and Berlin followed closely, each demonstrating thematic strengths in innovation and environmental performance. On the lower end of the ranking, cities such as Madrid, Warsaw, and Athens showed lagging performance, often due to weaknesses in either social cohesion or environmental sustainability indicators. To provide a more detailed interpretation of the results, a graphical indicator panel was created for each city. These panels visualized the relative performance levels of all 20 indicators and were color-coded based on score categories. Indicators with high weight but low performance were flagged as 'priority intervention indicators.' For instance, Istanbul's low performance in urban density and green space availability, despite their high weights, signaled urgent need for strategic attention. Lisbon, on the other hand, displayed strong environmental outcomes but moderate social inclusion scores. The panels served not only as diagnostic tools but also as communication instruments to facilitate stakeholder engagement. To evaluate the robustness and reliability of the proposed model, a series of sensitivity analyses were performed. Two main dimensions were examined. First, the orness parameter within the Bon-OWA operator was systematically varied from 0.1 to 0.9 to simulate alternative aggregation preferences—ranging from pessimistic to optimistic weighting of indicator scores. This variation was critical to test whether the rankings remained stable when decision-makers emphasized weaker or stronger performance areas. The results revealed that while minor shifts occurred in mid-ranked cities, the overall ranking order—particularly for top and bottom performers—remained consistent across orness values. This demonstrated that the model's output is not excessively sensitive to aggregation attitudes, supporting its robustness across different strategic preferences. Second, the linguistic evaluation scale used for expert judgments was altered. The initial 7-point scale was replaced with a more concise 5-point scale, and the entire evaluation process was repeated. This test aimed to examine whether a reduction in scale granularity would materially impact belief distributions and final city scores. The findings indicated that although individual indicator scores became less differentiated under the coarser scale, the overall city rankings remained largely intact. These analyses confirmed that the model maintains both numerical and ordinal stability under different evaluation conditions, making it suitable for application in real-world policy settings where data precision and expert availability may vary. To contextualize the proposed approach's performance and validate its added value, comparative analyses were conducted using established MCDM techniques and global benchmarking indices. City rankings generated through TOPSIS and SAW—two commonly used MCDM methods— and regular CBD were compared to those obtained through the proposed CBD-based approach. While high-performing cities such as Paris and Helsinki maintained similar positions across methods, noticeable deviations were observed in mid- and lower-tier cities. These differences were attributed to the compensatory nature of SAW and the geometric bias inherent in TOPSIS, which tend to overlook indicator importance weights and interdependencies. In contrast, the proposed approach's integration of expert uncertainty, non-linear aggregation, and thematic weight sensitivity allowed for a more contextually grounded and analytically nuanced ranking system. Further validation was performed by comparing the SSC scores from this study to rankings in global smart and sustainable city indices, including the IMD Smart City Index and the Arcadis Sustainable Cities Index. While certain overlaps were observed in the identification of high-performing cities, several differences emerged due to methodological distinctions—particularly in indicator selection, data sourcing, and aggregation logic. Unlike global indices that often rely on secondary data or perception-based surveys, the proposed model is designed to be adaptable, transparent, and locally calibrated. It empowers decision-makers to customize indicator sets, incorporate expert insight, and simulate multiple policy scenarios, thereby enhancing the practical utility of SSC assessments. In conclusion, this thesis presents a novel, multi-dimensional, and operationally flexible methodology for evaluating Smart Sustainable City Performance. It makes three key contributions: (1) the construction of a refined indicator framework grounded in systematic literature analysis; (2) the design of a robust, hybrid decision-making model that integrates Entropy, MEREC, CBD, and Bon-OWA methods; and (3) the empirical application and validation of the model across 15 European cities. Together, these contributions offer both theoretical insights and practical tools for city planners, urban researchers, and policy institutions. As urban challenges continue to evolve in complexity, models such as the one proposed in this study can support more informed, transparent, and accountable decision-making toward the creation of smarter and more sustainable urban futures.

Benzer Tezler

  1. Tez No

    962042

    Contributions to the determination of optimized driving strategies for electric vehicles using artificial intelligence based methods

    Elektrikli araçlar için yapay zeka tabanlı yöntemlerle en uygunlaştırılmış sürüş stratejilerinin belirlenmesine katkılar

    UFUK BOLAT

    Yüksek Lisans

    İngilizce

    İngilizce

    2025

    Elektrik ve Elektronik Mühendisliğiİstanbul Teknik Üniversitesi

    Elektrik Mühendisliği Ana Bilim Dalı

    DOÇ. DR. DERYA AHMET KOCABAŞ

    DOÇ. DR. GÜLCİHAN ÖZDEMİR

  2. Tez No

    665454

    A study of applying smart city standards on the city of Eskişehir

    Akıllı şehir standartlarının Eskişehir ilinde uygulanmasına yönelik bir çalışma

    FADI M. M. ALJERJAWI

    Yüksek Lisans

    İngilizce

    İngilizce

    2021

    Bilgisayar Mühendisliği Bilimleri-Bilgisayar ve KontrolEskişehir Teknik Üniversitesi

    Bilgisayar Mühendisliği Ana Bilim Dalı

    PROF. DR. CİHAN KALELİ

  3. Tez No

    888546

    Cultural heritage circular governance for the integrated conservation of historic cities: The case of the Medina of Munis

    Tarihi şehirlerin bütünleşik korunmasına yönelik kültürel miras'ın döngüsel yönetişimi: Tunus Medine örneği

    YASMINE TIRA

    Doktora

    İngilizce

    İngilizce

    2024

    Şehircilik ve Bölge Planlamaİstanbul Teknik Üniversitesi

    Mimarlık Ana Bilim Dalı

    PROF. DR. HANDAN TÜRKOĞLU

  4. Tez No

    932230

    Dijital ikiz teknolojisinin şehir planlama alanındaki uygulamaları: İstanbul Tarihi Yarımada Süleymaniye Bölgesi'nde kent silueti analizi

    Applications of digital twin technology in urbanplanning: Investigation of the historical silhouette inthe Suleymaniye District through 3D modeling

    SEVENAY UZUN ALTUNTAŞ

    Yüksek Lisans

    Türkçe

    Türkçe

    2025

    Şehircilik ve Bölge Planlamaİstanbul Teknik Üniversitesi

    Bilişim Uygulamaları Ana Bilim Dalı

    DOÇ. DR. ADALET DERVİŞOĞLU

  5. Tez No

    887125

    Enabling adaptive road lighting through lighting class prediction with real time and historical data

    Gerçek zamanlı ve geçmiş veriye dayanan aydınlatma sınıfı tahmini ile uyarlanabilir yol aydınlatmasının sağlanması

    HASAN MERT TOKGÖZ

    Yüksek Lisans

    İngilizce

    İngilizce

    2024

    Enerjiİstanbul Teknik Üniversitesi

    Enerji Bilimi ve Teknolojileri Ana Bilim Dalı

    DR. ÖĞR. ÜYESİ MUSTAFA BERKER YURTSEVEN

Geri Dön