Submitted by Berrin Chatzi Chousein

"The Path to more Sustainable Cities lies in Urbanizing AI" says Laura Narvaez

France Architecture News - Aug 28, 2024 - 10:25   1960 views

Laura Narvaez Zertuche, a leading advocate for Urban AI, envisions a future where Urban AI becomes integral to real-world projects for urban developers and stakeholders. 

As an architect, urban designer, spatial scientist, and advisor, she serves on the Board of Directors of Urban AI, a Paris-based think tank dedicated to ethical governance and the sustainable application of AI in cities. This organization focuses on the burgeoning field of "Urban Artificial Intelligence," aiming to build a comprehensive knowledge base by uniting researchers, public servants, startups, and urban experts globally.

Urban AI's core activities revolve around four main areas: research, education, knowledge sharing, and consultancy services. These efforts are designed to equip stakeholders in the urban sector with the necessary knowledge and strategies for implementing urban AI.

Laura Narvaez is also a Partner in the Urban Design team at Foster + Partners, where she has led data-driven urban design and masterplanning since 2015. Her approach combines evidence-based design with simulation-powered tools and strategic methodologies in architecture and urban design. Passionate about data analytics, Laura has volunteered as an advisor for Urban AI for the past three years. One of her significant contributions is the development and publication of the Urban AI Guide, which addresses the taxonomy of AI in urban contexts. This guide aims to bridge the knowledge gap between city leaders and the implementation of AI technologies in urban projects.

Urban AI Guide helps to contribute to a “knowledge gap”, says Narvaez 

Narvaez explains that the Urban AI Guide was created to introduce a new framework for understanding Urban AI. She collaborated with colleagues Sarah Polpeka and Hubert Beroche to develop this guide, which helps city leaders who often feel unprepared to engage with AI solutions for urban challenges. The guide dissects the components of Urban AI into what they call an "Urban AI anatomy," detailing the steps and considerations for implementing AI projects in cities.

In an exclusive interview with the World Architecture Community, Narvaez discussed three case studies from Winnipeg, Los Angeles, and Vilnius featured in the Urban AI Guide. These examples demonstrate how AI can provide effective infrastructure solutions on a city scale.

According to Narvaez, "city leaders who encountered new technologies as potential solutions for urban challenges. We found that they often felt they lacked the necessary background knowledge to effectively engage with and assess these proposed solutions with new technologies, especially with AI."

Urban AI Guide, authored by Sarah Popelka, Laura Narvaez Zertuche, and Hubert Beroche, proposes a framework for the implementation of AI into real projects. Image © Urban AI

She added: "We found that there was a “knowledge gap” between knowing the technology needed and its implementation in a project. Therefore, we proposed to develop a framework that exposes or dissects the components of an urban AI to what we call an “Urban AI anatomy. This meant understanding the steps and the considerations that go into implementing an AI project in cities."

"Technology shouldn’t dictate our behavior with the world"

Narvaez also highlighted the distinction between "smart cities" and the use of AI. She noted that the concept of smart cities, which aimed to optimize cities like machines, has often failed due to neglecting regional specificities, citizens' preferences, and the use of biased technologies.

"We need to understand first how we have defined the “Smart City”. For the past decade, the industry has promoted this concept with the idea of optimizing cities as if they were machines or computers," she said.

"I think this is proving ineffective or not being very successful. The reasons for these failures can be many: Neglecting regional specificities and citizens’ aspirations, ecological missteps, or the deployment of biased technologies that exacerbated urban inequalities."

She emphasized that "technology should not dictate our behavior with the world, but should honor our social contracts, enrich our heritage, and preserve our environment. In essence, it should help us inhabit the world in a much more sustainable manner."

Urban AI Anatomy by Polpeka, Narvaez and Beroche. Image © Urban AI

"We also want to assist in developing urban AI policies"

Urban AI's framework includes knowledge dissemination, education, research, and consultancy. Recently, the organization has begun offering consultancy services to government bodies and stakeholders to develop AI policies and urban design strategies. They also focus on climate change, organizing research and debates, such as AI for Low Carbon Cities, and capacity building to help organizations understand urban AI fundamentals.

"The first priority is to assist with urban AI strategies across various urban sectors. We also want to assist in developing urban AI policies. We see that climate change is also an extremely necessary subject to address," she said.

"There has been already research and debates that we have organized around AI for Low Carbon Cities - the article series developed by Urban AI, and we want to extend that also to urban stakeholders. And there is also capacity building, which essentially is to assist organizations with the foundations of urban AI," she added.

Research initiatives by Urban AI include the 91-page Urban AI Guide, various webinars and events, and The Screenless Cities Program, which aims to use urban environments as interfaces for AI, replacing traditional screens. Since 2022, along with the long list of contributors of Urban AI’s community, the organization includes a Scientific Committee, composed by Saskia Sassen from Columbia University amongst many others, to help launch the Emerging Leaders Program, - an educational program that aims to support young scholars and professionals in the field of Urban Artificial Intelligence.

"Overall, I believe that the path to more sustainable cities and the adoption of new technologies lies in urbanizing these technologies. By doing so, we can create more operational urban AIs, which are essential for the sustainable cities of the future," Narvaez concluded. 

Read on for an edited transcript of the interview:


Berrin Chatzi Chousein: Tell me what is Urban AI. Why did you decide to be part of Urban AI?

Laura Narvaez: I can start by telling you how it all began. In 2021, after exploring twelve cities and engaging with over 130 experts in technological innovation, a dear friend and colleague of mine, Hubert Beroche, established Urban AI. This interdisciplinary working group delves into the interplay between Artificial Intelligence and urban phenomena. It became or started to be established as a think-tank. It is a think-tank that wants to “urbanize” Artificial Intelligence. What is unique about Urban AI is that it is backed by a growing community of international experts in urban AI all over the world. Currently more than 150 contributors (myself being among them). Collectively, we are engaged in a variety of projects, yet we all share a common theme: A focus on Urban AI.

As a think tank and along with the +150 members, this led to an effort to create a Manifesto (Call to Urbanize Technology (urbanai.fr) at redefining the smart city concept, prioritizing people and their needs. We envision technology’s primary role as adapting to urban environments and developing new tools for active citizenship. The Urban AI Manifesto asserts that the only boundary for urban technologies is our imagination.

I joined Hubert on this journey in 2021 after learning about it. What started as a simple exchange of ideas soon evolved into collaborative efforts in educational programs, articles, and co-hosting events. I am one of the many contributors and advisors, and now acting as one of the Board of Directors, more like a “champion” (as we like to call it), to initiate, develop, disseminate knowledge, and innovate together -to keep the journey going and reach as much wider audience to join us in this community of Urban AI experts.

My primary focus within Urban AI has been on architecture and urban planning, which aligns with my professional background. With extensive experience in urbanism, spatial science, data analytics, and evidence-based planning, I’ve always been fascinated by AI. I’m particularly interested in how AI can be integrated into the future of urban planning and design to leverage advanced technologies for creating smarter, more efficient, and sustainable cities.

So, what is Urban AI? Urban AI is a Paris-based think tank focused on the emerging field of Urban Artificial Intelligence. It aims to create a comprehensive body of knowledge on urban AI by bringing together a global community of researchers, public servants, start-ups, and urban experts who work at the intersection of cities and technology. Through multidisciplinary projects, we strive to better understand and evaluate the impacts of artificial intelligence on urban life and vice versa.

The think tank has over 150 members from all over the world. Image © Urban AI

Berrin Chatzi Chousein: What is the difference between AI and Urban AI?

Laura Narvaez: The difference lies in their scope. AI is a broad field that involves the development of machines and systems capable of performing tasks that typically require human intelligence. These tasks include learning, reasoning, problem-solving, perception, and language understanding. It is a general technology applicable across many fields of disciplines. Urban AI addresses the unique challenges and opportunities within urban settings. It focuses on the application of AI technologies within urban environments, but more specifically, it refers to any system that gathers data from the urban environment, processes it using algorithms, and generates results that are beneficial for the socio-spatial aspects of the city. 

For example, with the growing challenges that cities face today and the desire of wanting to achieve more with less, AI has emerged as a promising solution. AI, defined as machines’ ability to perform cognitive functions primarily through learning algorithms since 2012, was initially integrated into large-scale so-called “Smart City” projects. Over time, AI applications in cities have expanded, driven by different stakeholders. At the moment, cities host numerous AI systems, owned and utilized by multiple entities to serve diverse, and sometimes conflicting, interests. So, the diversification of AIs in cities has become quite important. The question is, how is it used in cities? How do cities “respond back” to such technologies? Every city poses its own challenges, its own needs, and its own vision of growth. Every city uses AI differently. Therefore, AIs are also governed, implemented, and developed differently depending on policies, and culture. The term “urban AI” stems from those questions - due to the diversified range of AI applications in cities, which components unify those categories of AIs that can be used for possible solutions to their own needs and that can be deployed in its urban context.

Berrin Chatzi Chousein: There is a statement on the website: "Instead of creating 'Smart Cities', our goal is to urbanize Artificial Intelligence." What is the distinction between creating "smart cities" and utilizing AI?

Laura Narvaez: We need to understand first how we have defined the “Smart City”. For the past decade, the industry has promoted this concept with the idea of optimizing cities as if they were machines or computers. I think this is proving ineffective or not being very successful. The reasons for these failures can be many: Neglecting regional specificities and citizens’ aspirations, ecological missteps, or the deployment of biased technologies that exacerbated urban inequalities. One example to learn from this is the Sidewalk Labs project in Toronto, and to some extent, this has resulted in widespread disillusionment and even distrust towards urban technologies.

From the perspective of Urban AI, we are not saying that the problem is the technology itself. We believe that technology shouldn’t dictate our behavior or relationship with the world. Instead, technology should honor our social contracts, enrich our heritage, and preserve our environment. In essence, it should help us inhabit the world in a much more sustainable manner. Therefore, the difference between the Smart City concept to our view is that we rather pose the question: Instead of making cities smart through computational paradigms, how can we integrate technology into urban environments by taking inspiration from how cities actually are? How do cities perform or function rather than how they should be? By doing so, we can start redefining what it really means to be “urban”.

Urban AI organized a variety of webinars and events on Urban Artificial Intelligence. Image © Urban AI

Berrin Chatzi Chousein: Can you talk about the applications/services of Urban AI? Which industries can benefit from Urban AI and for what purposes?

Laura Narvaez: The types of applications and industries are also linked to the wide community of  Urban AI members. This community of experts works on different disciplines related to urban AI, such as Architecture, Urban Planning, Mobility, Data Governance, Finance, Public Policy, Software Engineering, Data Science, Biodiversity, Urban Economics, Diversity, Equity and Inclusion (DEI), Digital Twins, Infrastructure Management, Civic Design, Climate Change, Community Engagement, to name a few. 

Collectively, we focus our expertise on three main streams of work in Urban AI: One is the dissemination of knowledge (or our knowledge hub) which is through events, either in-person or online, and we produce the content of shared knowledge and information about urban artificial intelligence through written blogs or any other social media platform. This allows to spark new debates and also offers a platform for the community to use to share their ideas, current projects, publications, etc. URBAN AI Conversation on Generative AI and Urban Design can be watched on YouTube.

The second is education. We want to help new generations to better understand what urban AI is. We have our Emerging Leaders Program which is aimed at educating and supporting young scholars and professionals, from any type of discipline, to learn and propose new approaches to urban technologies. This is done through the support of the members of Urban AI contributors as well as the Scientific Committee of Urban AI -Academic Professors and City Stakeholders.

At the same time, we also want to help city stakeholders, both private and public, with the better use and understanding of urban AI. We have done this through partnerships that we currently are working with, such as the Chair Entrepreneurship Territory Innovation of Sorbonne Business School, the Institut Français du Design, PCA-STREAM, the Centre of Innovative Cities from Singapore University of Technology and Design, amongst others.

The third one is research. We research to advance the field of urban AI and contribute to international discussions on the topic. There has been research published in which we worked with other collaborators, for example with Cornell Tech, on the areas of what is the future of Urban AI, the geopolitics of smart cities, AI Localism, and most recently on generative AI for cities.

Berrin Chatzi Chousein: What does Urban AI mean for Architects and Urban Designers? Can they contribute to Urban AI? Or, how can they benefit?

Laura Narvaez: Probably there are two aspects to this. One is productivity and the other is in the design process itself. Designers can enhance productivity by leveraging various AI tools to better realize projects or speed up certain tasks. AI can help discover new materials, explore structural possibilities, or find solutions that consider various performance factors, such as environmental or financial. The other aspect is rethinking the design process itself. With generative AI, more voices can be included in the design process. Therefore, it is not only architects or urban designers, it is other disciplines that engage in the design process, enabling deliberations with generative AI and other innovations that create a new approach to design. The design process is more collaborative, and data-driven, it is inclusive and ultimately aims to identify and implement sustainable practices in urban design.

Berrin Chatzi Chousein: The website features an Urban AI Guide and includes three real case projects in Canada, the USA, and Lithuania. How has artificial intelligence assisted in these projects?

Laura Narvaez: First of all, I can tell you about what the Urban AI Guide is so I can explain the three case studies. The Urban AI Guide was meant to bring to the fore a new framework of what Urban AI is. This is a guide that I did alongside my colleagues Sarah Polpeka and Hubert Beroche. We began doing an extensive literature review of the implications of AI in cities but also in-depth interviews (eleven to be exact) with worldwide city stakeholders. City leaders who encountered new technologies as potential solutions for urban challenges. We found that they often felt they lacked the necessary background knowledge to effectively engage with and assess these proposed solutions with new technologies, especially with AI. So, we found that there was a “knowledge gap” between knowing the technology needed and its implementation in a project. Therefore, we proposed to develop a framework that exposes or dissects the components of an urban AI to what we call an “Urban AI anatomy”. This meant understanding the steps and the considerations that go into implementing an AI project in cities. 

Beyond describing our framework, we wanted to show how this is applied in three case studies. Winnipeg in Canada was about water infrastructure. The stakeholders wanted to build out an internal AI team for their water meter lifecycle and demonstrate how municipalities can adopt a phased approach to project implementation while simultaneously building their capacity. Los Angeles, California was about curb digitization for urban planning. This was mostly a role of public-private partnership and highlights the necessity of tailoring the solution to the specific problem and its urban context. Vilnius in Lithuania showed how they used a platform AI service provider that can offer AI capabilities with minimal effort required from the city.

Urban AI Publications. Image © Urban AI

Berrin Chatzi Chousein: How does the organization collaborate with governmental bodies and legislators to implement urban AI policies?

Laura Narvaez: We recently set up a consulting service as part of Urban AI. I talked about three main streams of work earlier. We are now extending our expertise to serve consulting in four different areas in which we think we can help city stakeholders and organizations: The first priority is to assist with urban AI strategies across various urban sectors. We also want to assist in developing urban AI policies. We recognize that addressing climate change is extremely important. Research and debates on a new article series, AI for Low Carbon Cities, have already been conducted, and we aim to expand these efforts to include urban stakeholders. There is also capacity building, which essentially assists organizations with the foundations of urban AI. In general, the future of Urban AI looks very promising. I think the next step is to become more operational, put all this knowledge and research into practice, and start creating real projects with real change. So, in that way, our idea of consulting is to do that. To take it to the next level.

Berrin Chatzi Chousein: In urban planning, an AI data repository can help plan a well-defined budget for its tools and integration into cities, which can expedite design processes. However, in this process, where citizen participation should begin and end in terms of data ethics is still debatable. Can you elaborate on this issue?

Laura Narvaez: I believe that citizen participation should start with data collection, ensuring transparency and inclusivity. Involving citizens early in the process helps ensure that the data collected reflects the needs and priorities of the community. On the other hand, there should also be ethical considerations during citizen participation. Citizens should be fully informed about the purpose of data collection and how their data will be used. Protecting the privacy of participants (data privacy) is crucial, and being inclusive. Representing diverse voices, especially those from marginalized communities. At the end of the process, I think, participation should continue through feedback and ongoing engagement to adapt plans based on the community’s input. The main challenges, I think, are really in balancing the benefits of AI with ethical use and addressing the digital divide - not everyone has equal access to digital tools or information and we should definitively be working towards a more equitable participation.

Berrin Chatzi Chousein: I've seen the title of Screenless Cities on the website. Can you elaborate on the Screenless Cities Program?

Laura Narvaez: This is something that is still under development. It’s quite new still. The program envisions using urban environments as interfaces for AI and information, replacing traditional screens. By leveraging elements like streets, public spaces, and buildings, the program aims to make AI and data tangible and interactive within the cityscape. We are trying to bring together innovative municipalities and urban AI experts to create pioneering, low-tech, multisensory interfaces that integrate AI into public spaces. We will eventually share the progress and the development of this program through the website, publications, and also public events soon.

Berrin Chatzi Chousein: Assuming that urban designers have access to a wide range of AI-powered advanced tools, and they are accessible by others too. Do you think developers will be able to directly handle design processes with these tools in the future, possibly without as much reliance on urban designers?

Laura Narvaez: I don’t think entirely without reliance on designers. I think the dialogue and certainly the collaborations are key. AI-powered tools are transforming our field, particularly in urban design and planning, for example by automating tasks and providing detailed insights in a much faster way. This could enable developers to handle some design processes independently. However, urban designers will still be crucial for complex problem-solving, creative input, and also for ethical oversight. The future, to me, involves a collaborative approach where both developers and urban designers, and even other specialists involved, leverage AI tools to enhance efficiency and creativity.


Top image in the article: Laura Narvaez. Image © Bernardo Borghetti.

This interview has been published as part of the Archi-ficial Intelligence series. 

We invite you to join our interview series to discuss the future of artificial intelligence together. If you're an AI-driven organization or have an AI-powered tool or service for Design, Architecture, and Urbanism, contact us to take part in the series: [email protected]. Selection for the interview at WAC's discretion.

AI Archi-ficial Intelligence artificial intelligence Laura Narvaez smart city technology Urban AI