Computational Urbanism For Informal Settlements: a Generative Design Approach In Këneta, Durrës

This study examines the integration and the potential of Generative Design (GD) as a novel, transformative tool in urban planning, and highlights its ability to improve decision-making to create more optimised, sustainable urban environments. Through a detailed analysis of specific key case studies, including projects that focus on the development of smart cities and residential neighbourhoods, the paper explores how computational tools --- such as Generative Design --- can optimise and automate the design process while balancing and addressing complex objectives, namely density, profitability, energy efficiency, and livability. The research employs a three-phase methodology. Firstly, it aims to create a theoretical framework that could define the concept of Generative Design by distinguishing GD from related computational approaches, including Parametric Design and Algorithmic Design, through its epistemological and methodological characteristics. Secondly, since informal settlements represent one of the most urgent urban challenges in rapidly developing regions, the area of Këneta, an informal settlement on the periphery of Durrës, is analysed through spatial, environmental, and social-urban investigations using a mixed-method approach that blends analogue mapping and computational tools like Grasshopper, Ladybug, and DeCodingSpaces. Thirdly, a series of design options is generated --- and further evaluated --- using the CityGenerator plug-in in Grasshopper™. The three proposed master plans emerged from a series of optimisations in the Galapagos engine and key criteria such as accessibility, density, visibility, solar exposure, and connectivity. The findings highlight the peculiarity of Generative Design to respect contextual constraints while, at the same time, accommodating and interpolating different and conflicting design constraints. The approach offers advantages in scenario exploration, design flexibility, data integration, and performance evaluation, attributes of particular value in informal urban contexts characterised by uncertainty and spatial heterogeneity. At the end, the authors conclude by highlighting several limitations, including the strong reliance on quantitative data that may overlook socio-cultural dimensions, and the potential inefficiency of design agency when computational tools are used without proper designer input and critical intervention.