SimAUD 2018 June 05-07 Delft, the Netherlands
© 2018 Society for Modeling & Simulation International (SCS)
Generative Urban Design: Integrating Financial and
Energy Goals for Automated Neighborhood Layout
Danil Nagy, Lorenzo Villaggi, and David Benjamin
The Living, an Autodesk Studio
New York, NY, USA
danil.nagy@autodesk.com
ABSTRACT
This paper demonstrates an application of Generative Design
to an urban scale through the design of a real-world
residential neighborhood development project in Alkmaar,
Netherlands. Problems in urban design can benefit greatly
from the Generative Design framework due to their
complexity and the presence of many stakeholders with
various and potentially conflicting demands. We
demonstrate this potential complexity by optimizing for two
important goals: the profitability of the project for the
developer and the potential for energy generation of solar
panels placed on the roofs of the buildings. This paper points
to further research into the application of the Generative
Design framework to solve design problems at an urban
scale.
Author Keywords
Generative Design, urban design, optimization, genetic
algorithm, parametric modelling, solar analysis, profit
optimization
ACM Classification Keywords
I.6.5 SIMULATION AND MODELING - Model
Development
1 INTRODUCTION
Generative Design allows designers to tap into the power of
computation to explore large design spaces and derive design
solutions which are both novel and high-performing relative
to a chosen set of goals. This process relies on a set of
technologies including parametric design software for
modeling the space of all possible solutions, simulation
software for deriving metrics to evaluate each potential
design, and optimization solvers such as the Genetic
Algorithm (GA) which can automatically search through the
design space to find the most optimal designs. In recent
years, this type of workflow has become widely used to solve
design problems in a variety of domains such as engineering,
industrial design, and architecture.
Urban design problems tend to be very complex, involving a
multitude of stakeholders, each with their own complex and
competing goals for the project. These complex goals are
difficult to resolve through a traditional design process,
forcing designers to rely on intuition and prior experience
which can limit potentials for novel design solutions. The
Generative Design methodology can help urban designers
navigate complex design spaces and a multitude of
competing goals across different stakeholder domains. Such
applications, however, have not been widely explored.
This paper describes a novel application of the Generative
Design methodology at an urban scale through the design of
a residential neighborhood of 7,000 sqm in Alkmaar,
Netherlands. To show the utility of this workflow, we
consider two important and competing goals, each
representing the desires of different stakeholders in the
project. The first is the cost and revenue of the development
project, which is important for the developer. The second is
the potential energy generation of solar panels attached to the
roofs of each building. This is important not only for
minimizing the environmental impact of the development
but also for the future homeowners who will benefit
financially from the energy being generated. Through this
example, we demonstrate how the Generative Design
process can help reveal the potential tradeoffs between
competing design goals and help urban designers discover
designs which solve these goals in novel ways.
2 LITERATURE REVIEW
While the application of evolutionary algorithms for
optimization of design spaces is well known within the
manufacturing industry [1], they are under-explored in the
architectural and urban design domains.
Prior work by the authors [2] explored the application of this
workflow to the interior layout of an office space. Calixto
and Celani [3] also described over 15 years of work
exploring applications of evolutionary computing for spatial
layouts. However, these studies were mostly theoretical and
did not show the feasibility of applying such a process to a
real-world design project. Furthermore, none of these studies
were applied at an urban scale nor dealt with profitability or
energy generation requirements.