digital urban planning

 digital urban planning

This research delves into the multifaceted realm of artificial intelligence (AI) applications within

architecture and urban design, shedding light on the intricacies inherent in contemporary design

processes. Focusing on the challenges faced in urban planning, the study advocates for a generative a

pproach to navigate these complexities. The paper delineates methodologies such as sensing analysis

and parametric modeling, incorporating urban strategies like Successive Palimpsest. Emphasizing

the pivotal role of digital tools and analysis in crafting well-informed design solutions, the research

showcases practical applications through case studies, such as BRNO.CZ - South Centre.

While architectural design tools have witnessed significant transformations over the past three

decades, the paper contends that their efficacy alone cannot address paradigm shifts or concept

development challenges. The key lies in the depth and direction of thought, shaping the subsequent

course of action. Since 2012, our research has implemented genetic algorithms and basic AI in urban

design modules, aspiring to forge an emergent dynamic model that not only facilitates adept tool

utilization but also introduces a novel urbanism concept befitting the 21st century. Termed "successive

palimpsest," this approach draws inspiration from natural processes of succession, fostering

harmony and climax in vegetation. The palimpsest, rooted in the scraping of Roman papyrus, a

dds a contextual layer, acknowledging and incorporating the wealth of information inherent in each

plot, place, and area. The term encapsulates an open yet defining system, contextual yet

devoid of romanticism, efficient yet intrinsically beautiful. The paper concludes by underscoring

the transformative potential of generative methods and encourages the exploration of future

possibilities in the realms of digital design and fabrication.

Egg or Chicken

Existing urban AI tools often gravitate towards utilizing established datasets or preconceived solutions,

functioning more as expedient modeling utilities than intricate generative models. A notable drawback is their

tendency to furnish solutions without adequately probing the essential questions critical for optimal outcomes.

While these tools may offer post-design analysis data, their effectiveness is hampered by the lack of a proactive

stance during the initial design phase.

In our innovative approach, we transcend these limitations by embracing a holistic utilization of all available

data. This methodology engenders a rich spectrum of potential solutions, meticulously cataloged into a user

-friendly repository. This unique cataloging system empowers users to select the most fitting solution aligned

with their preferences. What sets our system apart is its dynamic adaptability, actively learning from user

preferences. This adaptive learning process propels individuals towards realizing heightened levels of creativity

and originality in their design pursuits.

The intersection between Autodesk and OpenAI is pivotal in this paradigm. By leveraging the computational

prowess of Autodesk, known for its robust design and engineering software, and integrating the advanced AI

capabilities of OpenAI, our approach bridges the gap between conventional design methodologies and cutting-

edge generative AI. This synergy empowers designers with unparalleled tools, unlocking new dimensions

of creativity and efficiency in urban design.

Methodology

The foundational understanding of a city and the expectations we harbor from it should precede all

considerations, a realm untouched by the capabilities of AI. It is we, as humans, who must pose the essential

questions that define the very essence of urban living. Nevertheless, the realm of computational design

emerges as an invaluable ally, particularly in handling routine or information-intensive tasks within t

he urban planning domain.

Our approach initiates from the premise of envisioning a city as an emergent, pulsating entity intricately

interwoven with its surrounding landscape and nature. This conceptualization demands a substantial

influx of data encompassing terrain features, climate patterns, movement dynamics, program requirements,

and structural considerations. Leveraging sensing analysis, we extract this indispensable data, initializing

the generative process with the inaugural equation—termed as the 'vehicle.' This initiates the formation of

protomodels, subsequently applied to the designated site. Through a meticulous iterative process, we navigate

towards identifying the most optimal solution for a given urban challenge, facilitating a dynamic and responsiv

e approach to urban planning.