The Column: Laura Narvaez Zertuche on Urban AI

I think that integrating AI into the process of designing, planning and managing our cities has the potential to make them smarter, more sustainable, and better adapted to people’s needs. AI allows us to analyse and leverage vast data sets, such as population trends and traffic patterns, to optimise urban systems and resource allocation, while supporting planning and design decisions.

Generative design tools, powered by AI, can help us to optimise layouts for buildings, road networks, and green spaces, based on our chosen criteria. Then there are predictive analytics, which use machine learning models that can forecast population trends, urban growth, transportation demands, and climate impact. They help us to create resilient and sustainable urban plans that anticipate future needs.

We are integrating AI and generative design tools into our workflows to inform the urban design and planning process.

Our urban design team is using generative design tools that support the optimisation of urban layouts. These tools help us to produce 3D city models, which facilitate the visualisation and simulation of urban environments very quickly and efficiently.

We are also developing our own AI tools to meet our specific needs. AI workflows, such as machine learning models, assist in automating land-use configuration by analysing geospatial data, human mobility patterns, and social media inputs. This approach helps us to generate land-use plans, which align with our design workflows and the goals of our masterplan projects.

Everything we do is about creating spaces that resonate with people on a human level. We know that AI is a powerful tool, but it should never replace the creative judgement and expertise of the designer. Instead, it can be used to guide and augment the design process. AI’s output doesn’t always align with what we believe is right, so using AI requires critical thinking and a nuanced understanding of the urban environment to achieve a positive result.

Biased data is a major challenge that immediately comes to mind – and it can lead to inequitable outcomes.

To combat this, we always try to use diverse and representative data of the entire population or the specific area we’re looking at. We also try to use more accurate and consistent data collection methods to avoid measurement bias.

What’s more, we don’t only rely on quantitative data, but also integrate qualitative data. Our multi-disciplinary Urban Design group is made up of planners and spatial designers, economists and anthropologists, and landscape architects and mobility engineers. Our anthropologists and researchers collect qualitative data from interviews and surveys, which is then inputted into the design process.

AI analyses patterns in data, but high usage doesn’t always equate with desirability or preference. As I outlined in my previous answer, we must always consider the context behind the data – and combine quantitative data with qualitative inputs. This creates a more accurate picture of what people truly value in their urban environments.  

While AI offers incredible tools to shape cities of the future – I think it’s essential that these advancements remain grounded in human values.

A city is not just a collection of buildings and networks, it’s a shared space where lives and cultures intersect. AI should amplify our collective vision to foster wellbeing, celebrate diversity, and protect the planet – and help create an environment where everyone can thrive.