Ecological Urbanistic Analysis and its possible Implementation through Machine Learning - A Conceptual Framework for the Integration of Eco-Complexity into Analytical Tasks in Multi-Species Urban Design

Aug 3, 2026

Upload by Riley

Urbanization exerts pressure on natural environments within and outside urban areas, affecting human life and other species while threatening sustainable development goals. Rather than relying on a normative “green city” ideal, the thesis examines data-driven and performance-oriented approaches. It connects the ECOLOPES project’s urban classification for multi-species building design with a conceptual “ecological urbanistic analysis” framework. Using grounded theory and a review of theories and machine-learning methods, it evaluates potentials and trade-offs for implementing EUA via classification algorithms across spatial and temporal scales.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages199

Reading Time8 h 21 min

AccessFree to download

Abstract

Urbanization, sustainable development, and the need for data-driven ecology

ECOLOPES project and urban classification for multi-species design

Ecological urbanistic analysis (EUA) framework for complex systems

Methodology: grounded theory and literature review synthesis

Synthesis step 1: transect of theories for EUA

Synthesis step 2: review of machine learning methods in urban ecology

Findings: potentials, challenges, and trade-offs of machine learning implementation

Diplomarbeit (Diploma Thesis) Ecological Urbanistic Analysis and its possible Implementation through Machine Learning A Conceptual Framework for the Integration of Eco-Complexity into Analytical Tasks in Multi-Species Urban Design with Potentials and Challenges for its Implementation through Machine Learning Classiﬁcation ausgeführt zum Zwecke der Erlangung des akademischen Grades Diplom-Ingenieur / Diplom-Ingenieurin an der TU-Wien, Fakultät für Architektur und Raumplanung submitted in satisfaction of the requirements for the degree of Diplom-Ingenieur / Diplom-Ingenieurin at the TU Wien, Faculty of Architecture and Planning von Alexander Authried, BSc 00626359 Betreuer: Univ.Prof. Arch. Dipl.-Ing. Michael Ulrich Hensel, PhD Mitbetreuung: Assistant Prof. Milica Vujovic, PhDE259 Institut für Architekturwissenschaften E259-01 Digitale Architektur und Raumplanung Technische Universität Wien, Karlsplatz 13, 1040 Wien, Österreich Wien, am eigenhändige Unterschrift i Abstract Urbanization exerts pressure on natural environments within and outside urban areas. This aﬀects human life as well as other species and threatens the goals of sustainable development. While most attempts in architecture and urban design to cope with the challenges imposed on cities are rather following a normative ideal of the ’green city’, recent approaches are aiming at incorporating ecological knowledge through performance-oriented and data-driven design methods. The ’ECOLOPES’ project is developing a recommendation system for multi-species building design in urban contexts. Within this project, urban classiﬁcation is determined to identify potential project sites with similar urban and ecological conditions. Such classiﬁcations are a crucial analytical foundation to purposefully develop ecologically sophisticated design proposals. Although urban classiﬁcation has been a research topic with many applications, a conceptual approach to develop classiﬁcations as analytical method, addressing the complex behaviour of ecological systems, has not yet been undertaken. Machine learning provides a set of tools which are promising to engage in new ways with big data and its underlying patterns, but research in the context of ecological urban design is sparse. As machine learning studies are characteristically experimental and case-based, and theoretical foundations for interdisciplinary application are still lacking, this thesis sets out to conceptualize ’ecological urbanistic analysis’ (EUA) as a framework for analysing complex ecological urban systems at multiple spatio-temporal scales and further assess the potentials and challenges to facilitate the implementation of EUA in urban design through machine learning methods. This thesis is methodically based on grounded theory and uses elements of literature reviews to support the comprehensibility of the results. The main part is a two-step synthesis: 1) a transect of theories of biodiversity, ecological systems and spatial analysis as a conceptual framework for EUA in the context of multi-species urbanism, which incorporates diﬀerent dimensions of eco-complexity, and 2) an assessment of current machine learning methods and algorithms in the context of urban ecology with a short review of case studies. This thesis concludes that machine learning oﬀers a novel and interesting approach to analysing complex ecological systems within urban areas, where diﬀerent algorithms are suited fordiﬀerent aspects of eco-complexity. However, there are also important trade-oﬀs and challenges that come with machine learning as computational method. Although up to date only few studies at small scales, addressing issues of urban ecology and biodiversity, have been conducted, there is a lot of potential for future research, and ultimately a chance to mainstream ecological knowledge into urban design and architecture, to support long-term sustainable urban development. Keywords: urban analysis; urban ecol

FAQs

What problem does the thesis address regarding urbanization and sustainability?

Urbanization creates pressure on natural environments, affecting humans and other species and undermining sustainable development goals. It motivates approaches that integrate ecological knowledge beyond a purely “green city” ideal.

What role does the ECOLOPES project play in the thesis?

ECOLOPES develops a recommendation system for multi-species building design, where urban classification identifies potential sites with similar urban and ecological conditions. The thesis uses this context to argue for stronger analytical foundations.

How does the thesis propose to implement ecological urbanistic analysis using machine learning?

It conceptualizes EUA as a framework to analyze complex ecological urban systems at multiple spatio-temporal scales. It then assesses current machine learning methods for urban ecology to facilitate implementation through classification while considering trade-offs and challenges.

Aug 3, 2026

Upload by Riley