Enriko Lill

• Application of Generative Design Method in Analysis of Urban Planning
• Tutor: Raul Kalvo Martin Melioranski

The world is constantly changing, bringing new challenges and problems for developers, architects and planners. In order to create a sustainable, durable and livable environment, it is necessary to make increasingly complex design decisions, where numerous factors from different fields must be taken into account. In classical planning, there are many different complex stages between the completion of a sketch and the final project. Decisions made at an early stage have a direct impact on the further development of the project. Many decisions are made based on various analyses and simulations.

Often our work is limited by the lack of time resources, which does not allow us to analyse if what has been done is enough and to find compromises between different factors. The generative creative process makes it possible to quickly create many different models in the field of planning and to compare and analyse them at the same time. In this way, it is possible to find the best and most thought-out environment from different results.

Generative design is a method of applying computer-based performance to design. Instead of focusing on the conventional form-finding, the generative design focuses on describing the constraints and objectives of the design space.

The primary goal of the research was to explore the nature of generative design and methods of using it in planning and architecture. The research showed that in practice, the generative design can be applied in different stages and levels of the project to achieve the best design goals. The most important thing is to make the target parameters comprehensible to the computer and set the proper evaluation criteria that define the design goals. The computer does not generate new solutions on its own but works only in the framework given by the user. 

As a part of this master’s thesis, the author built a prototype of an environment for implementing the generative design in the urban planning process. The prototype consists of a parametric urban design model, the analysis of the model and pre-existing optimisation algorithms. In the course of the project, a toolbox for creating parametric designs was created in the Grasshopper environment to optimise complicated design situations. 

The new toolbox was applied to the area between Väike-Õismäe and Mustamäe. In the course of the work, three hypothetical urban planning scenarios were tested. Considering the analysis of the master’s thesis, the solutions that were aligned in the best possible way with the set vision and goals were selected from the results.