Day 2 - Integration and Strategies | Future Cities Symposium & Exhibition 2025

Okay, this is Stravko Trivic. This is me, Christoph Holger. We're talking about Architectural Cognition in practice as part of the Future Cities symposium. That's a lot of slides. Okay, now some content. A core content of what we do is driven by our research team. I'm a psychologist by training, Stavko is an architect by training. The third person at professorial level who serves at the core team is Panos Mavros. He used to be the module coordinator for us over a period of almost seven years, and then he got hired into Paris Telecom where he is connecting human computer interaction, design and architectural topics now. And the other members of the research team? Leonel and Rafael are the members of the team back in Zurich. They're not here today, while Lara, Gregorian, Chuchung, Freyan and Azrin are all in the room today. And if specific questions come up later that pertain to their respective work, they of course are more than welcome to answer instead of the professors.

Clearly, at the core of what we do is interdisciplinarity. That's represented by the home disciplines of the two of us, but it's even further represented really in the composition of the teacher. We have two computer scientists, we have three architects, a designer and a neuroscientist among the researchers in the direct team. And we also draw both at NUS and at ETH Zurich, then from resources of other architects, urban planners, psychologists, neuroscientists and human geographers.

What are we talking about? We're talking about the built environment.

And we.

And more specifically, we're talking about the human in the built environment. And these three images. Here are examples of what we have been done and what we are increasingly doing. In the beginning of our collaboration, we mostly worked in the public realm, in complex morphologies of buildings with a lot of foot traffic, lots of pedestrians, that could be transport stations, airports, and in the case of Singapore, predominantly those mixed use scenarios that we often find in and around shopping malls, where you have atria and pathways and lots of buzz and people. Another work stream that we opened up to more in the last few years is looking at more stationary activities, not so much pedestrianization or moving about like you, for example, find in office spaces, where you then have questions that are maybe less under the label of livability and more under the label of health and well being. Stress, task oriented performance, attention, being able to concentrate in the workplace, having good communication structures. That's something we are increasingly looking into. And one scenario, one type of building where these two things come Together is actually the hospital. The hospital is often in environmental and architectural psychology used as a space of either very stationary things, you are a patient or you are a nurse taking care of a patient at a specific point, or as something that's about how you move about as a visitor trying to find your way in the maze of a hospital. But in our recent work, as I'll introduce a few slides down the road, we bring these two things together with a workplace oriented question. So let me talk about our role a bit. Right, So I already introduced the user of buildings as a key stakeholder. While in many of the other modules we quite naturally always talk about stakeholders such as the designers, the people who are responsible for the spaces. And for a long time we've worked with this triangle logic. There's the designers that design buildings, there's the users that use buildings. And then the crucial relation is the third relation. The designer needs to anticipate the user, the user's needs, the user's wants, the user's tasks, the user's abilities. And how does one do that? Traditionally that is largely driven by tacit knowledge, by intuition, by the kinds of things that you pick up in your education and professional experience. And we do believe that, and that is there's a tradition there of environmental and architectural psychology and human centric evidence based design and architecture that looking at this from a human centric psychology, cognition driven perspective can help in two ways, via the pathway of science, namely looking at peoples and users in place and over time to understand the cognitive processes and how the built environment influences the user and distill from that knowledge about informing the designers. To get even better at this anticipatory relation between design and usage. And without going into all the details of our program, that's definitely outside of the scope of our 10 minute slot. I would like to introduce that over the last few years we've developed this three pillared framework where we differentiate between fundamental research, reflective research and translational research. And these three streams of research clearly influence one another and inform one another. Thus all these arrows. The fundamental research looks at how do people behave in the built environment, how do they perceive it? In scenarios like those that I have introduced, the reflective research looks more from a design studies, design research point of view at how does the designer think about users, how does the designer know about users, what assumptions do they have and how do they translate that into design decisions? And then the third pillar is translational research that tries to take the outcomes of these two other research streams into practical solutions, into guidelines into card decks, into interventions, into teaching activities and best practices that could then be taken up by industry. And luckily over the course of the years, we started out largely with doing fundamental research only. But step by step we are bringing on in various domains industry partners that increasingly listen to what we have to say and increasingly also co create data to get a better understanding of what their respective projects require.

And I'll just talk about one example that is currently only driven from our work at the Zurich hub and then I'll give over to Zdravko. But it's not only happening in Zurich. The two slides I'm talking about here are examples from Zurich University Hospital where we have an ongoing collaboration with the emergency department trying to understand how the spatial arrangement of the emergency ward, how that influences their work processes, their ability to obtain the required information for their tasks by communicating with one another, having access to the patients and other factors. We do spatial analysis. Thomas Schepfer earlier today spoke about using techniques such as space syntax at the building and also at the district scale level. We are using this really within much smaller spaces to investigate what are movement routes, what are opportunities for visibility, interaction and communication. And we relate those space index analyses to movement patterns. And for doing that in a privacy consistent, ethically approved way, we instrument volunteer participants with body worn sensors, basically RFID tags. We instrument the space with sensors and collect voluntary movement traces of staff members with different roles and relate that to the spatial analysis and find a very good correspondence between space and text predictions and locations of communication, locations of individual work, locations where there's more positive interactions and locations where more interruptions are being perceived. While this is observational work in a hospital, and work in a hospital is very difficult to do in an intervention way because in an emergency ward there are lives at stake at real time. We work with the Swiss center.

With the Swiss center for Design and Health in Nidhau and Canton Bern and they have a huge projection space where you have a top down projection where you can, together with physical props, build up mock hospital wards and can compare different role playing scenarios and look at how a change in spatial arrangements changes communications patterns and the patterns of interactions in such a role playing scenario. And we've recently teamed up with HKS architects in Singapore who are doing similar simulation scenarios with different, different technologies. And we're increasingly exchanging about how in the future we could work together on improving hospital design with that. That's a working mic. It's a good mic.

Even better.

Thank you very much, Christophe.

All right, so among the projects that we have completed throughout this module is one of the fundamental research that I did about using multi sensory approach to really understand how older people in their familiar HDB environments perceive and interact with spaces around them. This was one of the first studies, if not the first as I know, that used eye tracking, mobile eye tracking in the real world environment with older people in Singapore to really understand how the things work for them in their neighborhood, in their familiar environment. What was done here is we used, we implemented surveys together with mobile eye tracking and postwalk interviews to really understand the qualitative data behind the visual attention. So the idea really here was to use eye trackers as something that is not distractive, but rather puts attention to other sensory modalities so that we can understand the spaces in a more alternative way and provide alternative measures for aging friendly or aging ready ready spaces. What is important is that when we link these kinds of researches and different types of engaging stakeholders. This was one of those studies that we do in school. We do it in studio with students. So it's more like that the studio informed the research and the methodology in this work rather than the other way around.

Another project that is done by our PhD student Freyan is a small pilot study whereby she tried to again use eye trackers, but to rather understand what is the relationship between the spatial complexity in the environment and place, or rather place making. What she did is that she used the models of five complex shopping mall or rather mixed use developments around the world. One of them was from Singapore. She did it in a small study of 20 people. They looked into the videos of very bare models of these spaces and she tried to really understand whether even such an environment without any other kind of information in it can trigger certain behaviors or certain attachments to certain points in that space. And what was really interesting is that some of the statistically significant results actually showed the relationship between the level of complexity and.

The attachment to certain parts of the spaces. What she did is one of the questions, for instance, how to intervene or rather understand the identity within such a very bare space was where would you put an art piece in this space? And people were resonating with this question and they tried to put meanings into this kind of environment.

In another study that was done by our PhD student Yu Qin, we also reviewed the state of the art in the literature. What are the tools and methods that are human centric and used in the different kinds of research and practice and that is already published. What Yujin did is that she created a very hierarchical systematical taxonomy of different methods and tools. And she also tried to find the links.

Between them and the human centric approaches to design in such a way we didn't only review the field in a reflective way, but also created a kind of a knowledge base for the practitioners to use in their design process.

We went a little bit further as well. This is something that we started with from the very, very beginning, but it ended up being done relatively recently, whereby we engaged people in practice, built environment practice, to really understand to what extent they really engage with the ideas of users of human centric design. Or how can we use anything related to evidence based design or the spatial cognition and architectural cognition within the design processes. And we had this with more than 150 participants in almost 40 countries around the world. And obviously some of the key findings here are a little bit biased in the sense that people who responded to us are those who actually are involved and want to use this kind of methods in their work. But what was really interesting to us is that out of all the things that they were really interested in or the kind of importance for them was really the human centered design, whereby the evidence based design and spatial cognition was something that was a little bit less known and a little bit less utilized in their design process.

It is also important to highlight that.

Among the human centric topics, they were really particularly interested in the ideas of psychological and emotional well being, cognitive processing and decision making, user experience and satisfaction and social interaction. But one of the challenges, the key challenges to really bring in some of these methods and knowledge into the design processes were really time and resource commitment and constraints, a lack of interest from the client's point of view, which is really sad, because the clients do not make necessarily only for themselves, they make for other people. But also the difficulty to access data and information, but also to integrate the research kind of intensive kind of methodologies within the design process.

What we did next is we tried to really engage with different kinds of stakeholders in so many different ways. We went to different conferences, we engaged with people formally and informally. One of the formal ways to do so is that we did different kinds of workshops with practitioners, with agencies on the spot. And this was really an important way for us to, beyond the survey and beyond that kind of outreach, also get the important information.

How they utilize.

Human centric methodologies in their design practice.

What we also did is that.

Eugene also developed a framework for evidence based design in context of another very important way of engaging with the stakeholders. This was research, or rather practice at Convent, whereby Eugene spent some time in the UN Studio in Amsterdam to really work with them on a project for the light rail transit station design whereby she engaged with the practitioners to develop a methodological framework, but also to refine her own framework and to develop different ways to help their decision making and brainstorming processes. So what she did is she did different kinds of spatial analysis, VR validation experiments to really not only inform the designers to make their design better, but but also to revise and refine her own methodology and evaluation of other designs.

This is the last project in what I call the laundry list of the baby steps that we did in engaging the different stakeholders. This particular project was led by Azrin and Panos. It really started with the initiation of LTA to create better way finding systems in different types of environments. One in residential, the other one in cbd, more commercial and office development. And it didn't stop there. It didn't only help to create the signage system, but also to engage the end users and then re evaluate the design and then reassess the possibilities to improve them in another way. So what was done again here? The users used mobile app trackers and interviews, walk along interviews to really assess the accessibility and the values and success of these wayfinding systems in these two environments.

And in the end, this is our last slide. We just wanted to give a glimpse of the future. As I just mentioned, we feel that what we did here is really like a series of baby steps with interacting with different kinds of stakeholders. It's messy and we are not really afraid of messiness. In a way we are embracing it. But in the next step we really want to put people at the center of it. We are looking at places for people and places with people, I would say, whereby we really try to look into qualitative and experiential aspects of human experience and then how people use spaces, how they appreciate them, how they negotiate and appropriate different kinds of dynamic environments, but also through the lens of cognitive and behavior, more empirical evidence based lenses and also with assistance of predictive new AI technologies that serve not to give us solutions, but rather to kind of help us in sense making of the potential quantitative and qualitative data that we gather throughout these processes. I would like to thank again to the entire team for making the.

Making these outcomes happen. Thank you very much.

I will be presenting two of our work packages showing you some research we have done trying to understand the welfare that people achieve by living as residents in a city and bringing a perspective of resource efficiency. So how can we make the best use of the resources we have available to support the lives of people.

Okay, yeah. So Work Package 1.

Is headed by Lok Yu Chu at NTU and Marcus Schleppfer currently at Columbia with Tianyu as the postdoc, working together on analyzing empirically social interactions in Singapore and in particular bringing in a efficiency perspective. So in the figure two to your left, you will see a heat map of social interactions proxied by spatial interactions. So how often do we see two people staying at the same location for a certain amount of time and what's the spatial distribution here? So this is the current state of Singapore in the mobility data that we're using. And then what they're interested in is to see how do we support these interactions in terms of the necessary travel time that you have to take as a cost in order to get this important utility of living in the city. So that's their first set of analysis. So they did look at how many interactions do we get per unit traveled. So a distance perspective in terms of the cost and what we can see on the figure on the right hand side is that there are places that are fairly inefficient where you typically travel quite long to go there. So for example, gardens by the bay, if you observe two people interacting there, you would expect that both of them had to travel quite far to be there, while there are other places.

Where you are typically much more efficient. So a key contribution here is to map out spatially this distribution and then take it one step further, asking what can we do to potentially allow such interactions in a more efficient way? So what work package one did here is to say, okay, these are individual trips. So we considered each interaction that we observe in the data an individual trip. And let's say if you can bundle them, what happens to the efficiency measures? So they took all trips. So every interaction, for every spatial interaction they looked at, if it happens together with another.

Set of interacting agents that travel at the same time from the same origin to the same destination, then you just bundle those trips, put them together. So then you have fewer, in some sense, you have less travel cost for the given amount of interaction. So that's what's going on from the figure three on the left to figure four on the right. And we do see that there is substantial scope for improvement through bundling in the transportation network. So this is useful as a tool, an input to policy.

By seeing where are the largest gains, so where's the largest scope to start improving transportation efficiency? And naturally, many places like the downtown core see a large, large potential improvement. The Reason being, many people like to travel there. So there's a lot of potential.

Demand for these type of bundling, so services. And then they take this further and look at, okay, what's the distribution across industries and these type of analysis. In the interest of time, I will not go much into detail there, but we can mention both substantial improvements in terms of residential and business. So that's where you see a large.

Push from the old to the new.

Equilibrium in terms of this simulation. Okay, so that's from a working paper of work package one. I'll move on to work package two, which I have been involved in together with the PIs in our module. Peter Egger at ETH Zurich and Xuan Chong at NTU. So we build on these type of patterns and we asked the questions looking at the travel patterns uncovered in Work Package 1 and additional information about characteristics of where people live and what's the wage distribution across Singapore. We want to know what are the preferences that underlies these decisions.

That by itself is interesting, but really the, the purpose of this analysis is to be able to characterize the optimization problems that generate this type of behavior. If we can do that, we can resolve them for different structures. So this is the tool we built for urban planning. So what this allows us to do is to say if we do an infrastructure project, maybe we improve amenities. As I will show some examples of, we can solve for what the data look like in the new hypothetical situation and then we can compare to the original. And that allows us to assess efficiency gains, for example. So I'll just briefly walk you through the model that we set up. So we take the perspective of the resident. A resident has several choices to make. The resident has to choose where to live, meeting a, a housing market. So at the other side there's a landlord. Both of these agents have preferences and they optimize. The resident also chooses where to work, earning a wage, meeting a labor market that has to clear, and so on. The agent also sources goods and services from throughout the city. So again, interacting with the firms, but from the demand side of goods instead of the supply side of labor. And finally, we added two new components to the literature. We let people travel around the city. So we have amenity visits like this heat map is showing. So people travel around just to enjoy their life in the city. And importantly, as was also documented by Work package one, we let transportation costs be endogenous. So if more people start to move into a certain area, that increases the cost of getting there. So cost is not just distance here, it is also the time you spend. All right, so we set up the model, and then we built a tool to solve this model. So this is basically every interaction you see here. Every market is a set of nonlinear equations. All of them have to be solved at the same time. So we spent quite some time developing this framework. This is just a schematic of how this solution works. I will not go into the details. The core idea is that you have to find a set of, for example, wages that are such that when I, as a resident, look at them, and when the firms look at them and we both decide where I should live and how many people the firm will hire, both of us have to make decisions that are such and consistent, so that when the effect of our decisions feed back into the new city structure, it is consistent with the input. So it's a fixed point problem. All right.

So just to give some intuition, this is from a simulated city. So here we can see this is the transportation network in the city. So it's fully connected.

And at the right you can see the productivity. So that is where is it most beneficial to locate businesses. So here there's a clear cluster. You would like to put your firm there, if you could. It's the best place to build goods and services. And then we consider various scenarios. So we look at four different worlds. So the rows, the top row has high transportation cost. It's very costly to move around the city. The low row.

Is cheaper. And then we do an experiment with amenities. So the first column has residential amenities that are very clustered, relatively close to what will be the business center due to productivity. And in the next column, we disperse amenities. So that's an amenity investment project. And we can see how the resident location choices are affected. And we can map out how the basically the heat map and the spatial distribution of residential location choices after the fact.

So this highlights a few mechanisms. If transportation costs are very high, you will see that once you go from a concentrated to a dispersed amenity structure, you actually end up with a fairly concentrated residence location. While if you are able to lower transportation costs at the same time as increasing the amenities, you get a more dispersed distribution of where people live. And that is because of the equilibrium effects. So this goes into wages and everything that was shown in the relations on the previous slides. So you would think, for example, in a world where you have high transportation costs and concentrated residential amenities, you might want to move. Everyone would want to live here right next to a business, right? Because it's very costly to travel. Why does that not happen? Well, there's an equilibrium effect. Once people start to move into that area, it pushes up real estate prices. It becomes very costly to stay there. And that pushes some people to move further away, depending on things like how much, how much they're making and other budget constraints. All right, so these are the mechanisms in a simulated city. What we're interested in is of course not simulated city with actual policy advice. So this is commuting patterns in Singapore. It shows the fraction. So where are people commuting to the downtown core basically sourced from? And we can see that it's, it's quite dispersed and there are a lot of people who commute quite far. So this is exactly the patterns we impose on our model. And we make our model explain these patterns. In other words, we characterize the preferences of people and firms that generate this type of pattern. And then we can start to do analysis. And that will be my final slide. So here I've done one experiment. What we did here is that we took three sub zones in Singapore and we increased amenities, residential amenities, by 30%. And we did this many, many times. So what happens is that every dot you see here is a different Singapore. So it's a different counterfactual simulation. And then you are able to compare them in terms of things we would like to have more of, such as welfare. Welfare. Here is the expected utility of living in Singapore, treating everyone the same. While the cost we are considering is traffic. So it shows there is actually an efficient frontier in terms of planning. You can, if you fix a certain. Does this one have a pointer?

Oh, yeah.

So if you fix a certain traffic level, you can have a range of utility levels, but if you fix the traffic level, you'll always go higher. But if you pick the highest utility level for that traffic combination, you could actually lower traffic and have the same level of utility. So eventually this feeds into an efficient frontier of the best possible policies if you care about welfare and traffic. And then our model has a lot of state variables. So as a tool you can do this type of efficiency trade offs for any variable you would care about as a planner.

All right, with that, I'll give the word to Felicity.

Hi, I'm Felicity and that's my team, the World Package three. This is a very different work package from the other two is really focus on more of the qualitative aspect of idea of interaction. So in our work package, instead of looking at interaction efficiency, we are looking at interaction effectiveness. So my work package, composed of myself and Patricia, who has now finished a contract with eth, but she's been somebody who has been with the project since day one and we continue to work together. And a summer intern, guy Bo, who is at ETH and a part timer and both Gaibu and Eugene just joined us in June. Let me get started. So, okay, so I think what I want to bring up here is really a conundrum that while cities are recognized as creators of innovation, and I recall what Professor Louis Betancourt said, what are cities for? Right. I think was it on Monday. And he said really cities are meant to a place for social interaction. With social interaction you could expect innovation to happen, you could expect agglomeration economies to be achieved, and so on and so forth. But we also know that being dense, large and diverse are also conditions to result in loneliness, anomaly, stimulus, overload, social fatigue. And this is well documented. So what do we do with this kind of findings? Is it good to densify how much and what does it do for our social interaction? So the purpose of this project is really something very practical. It's a practical question, but an aspirational one as well. How do we plan and design growing and densifying cities that offer social spatial intensification without compromising on the vitality of the cities? So let me just show you the next thing. So with this aspirational and practical question, how do we actually study this scientifically? So that's our question. How do urban residents perceive and experience urban vitality through public environments in these densifying cities? Comparing Zurich and Singapore and what can produce, what kind of environments can produce and sustain urban vitality in conditions of densification and intensification?

What did we do? We did an online survey. It would have been great if we have gone on site, but we do not have the manpower. So we did online survey with 1,281 respondents, over 900 in Singapore and 300 in Zurich. This was something that we collected about 1,900. This is just sharing the difficulties of an online survey. A lot of them were bots, right? So you have to clean it out. So by the time we did this is what we got. So we did it over two months, two to three months in both cities concurrently in English and German, and 36 questions about personal experiences and also demographic traits.

Well, so urban vitality, for us, we define it because we need to define it for people to then tell us what they think about this. We define it as life giving, engaging and restorative experiences. But what we did was we also asked people what Is your ideal vitality and why did we pick vitality instead of vibrancy? Because vitality has the word vitals, right? We all know that. What are your vitals? How are you doing? So there's an aspect of well being there but there's aspect of interaction which is at the heart of this project. So you can see from here people are huge deal people but from here you can't tell people it's like too many people, too few people, what kind of people and green is quiet, Green is what they think their ideal way of what urban vitality is.

So we have the question where are these urban? I mean where these vital and non vital places. We ask them to identify the places, right? So in Singapore what you can see is that they're all around. Many of them are places which actually double edged swards. Orchard Road is both vital and not vital. Where else? I mean if you're familiar with Singapore Sentosa, both vital and non vital. So there's quite a few of these double aged swards. But there are places where certainly the vitality is greater than the non vitality which is East coast park if you're familiar with it. So this is the picture of Orchard Road for better or for worse is both vital for some and not vital for others. So let's look at Zurich. Maybe it's not unexpected if you visit Zurich the Harbay the train station is both right, it's both vital and non vital. A double aged sword. But the lake area is generally has got a lot more vital points other places where like Kina Visa which is an area next to the lake where people go for picnics. When we ask them why is this like why do you consider this vital? And they tell us we can gather with our friends, we have picnics, all that. But similarly others, they are small group who think it's not right, it's noisy or whatsoever. So some of these places might come as a surprise. The Glad Centrum is a shopping mall. You might not think that this is the last place of vitality but there's a big group of people who actually like it and think that it is engaging, it is restorative and life giving. Right, next. So yeah this is a picture of the train station, of the main station in Zurich. We asked them who do you usually visit with? I think from this slide you could quickly see that in Singapore people hang out with each other a bit more than in Zurich and it's quite significantly so. So what does this mean for social interaction? Most likely if you're with your friends, you're not as open to meeting new people. Not all the time, but most likely if you're in a public environment, your friends are talking to your friends. So the kind of interaction is also expected to be of a certain type. Okay, this one. So we asked them, within 50 meters, how many people do you want around you? Because we want to understand the perception. And in Zurich.

Surprisingly or not, people could take more people. In Singapore, not exactly 50 is the maximum. Maybe it's because Singapore is just a bit more crowded and you cannot imagine more than that. Participants in Zurich have a higher people density threshold than those in Singapore.

We asked them how long they have lived in the city. From here you can see that in Zurich, the ones who have lived in a short time can take a lot more people within 50 meters than the ones who have lived there their whole life of more than 20 years. In Singapore, not exactly. I think maybe everyone is just feeling a little too cramming. It doesn't show too much of a difference. But overall, the longer term residents of cities tend to prefer fewer people for urban vitality. And I think a hypothesis here is possibly, or rather an explanation that we could think of here, is that if you have lived in the city for the longer time, you have also been in the city while it was smaller. And so your threshold of like, when, when has Singapore become so crowded? When has Zurich become so crowded? So you don't like as many people around as, you know, people who have been here chosen Singapore and Zurich because of the number of people. So it's a bit of a self sorting that way. Okay, this is also, we want to get really sense, right? It seems like people are a little finicky about how many people are around. So we ask the question, does it ever get too crowded or too few people? Because when it's too few, you also don't want to interact. I mean, the social interaction is not great, right? So apparently there's no statistical difference. But what we have is people who said yes to both and yes, no. And so we named them as the balanced believers. Say, like not too many, not too few. The Goldilocks, right? You know, the ones who just want your porridge the right temperature. But a good chunk of people said, yeah, it can get too crowded, but no, you know, it doesn't get too few. So I think they are looking for calm in the city, in both cities. And of course these two groups, the ones 8% are like more people, the more the merrier, right? And this one is like, I don't know, anything goes. Okay. So we also asked them, right, if you find it too crowded, what is it when it's too crowded for you? I think clearly in Singapore, if we just look at the mean, the highest mean here, let's see here. Too much queuing and waiting. Time is of the essence in this city. Everything needs to be efficient. If I get to wait too long, unless it's for food, maybe that's a tad too long. So this is a queue at Orchard Road waiting for an ice cream. Maybe that's okay on a Saturday, but in Zurich there's no space to sit.

So this is an experience when it gets too crowded in Zurich, so maybe there should be more public furniture, I don't know. But anyhow, these are what the results say. This is the last slide we have. Really? Then we ask them. So the next question is like, the planners and designers need to know, so what right with all this. So we asked them, what are this list of environmental features, which is for the kind of social interaction that gives you life giving, engaging and restorative experience. And for Singapore and Zurich, without the convergence is that clean grounds and surfaces with no litter, presence of trees and shrubs.

For Singapore only, no smoke or bad smell in the air, feeling of openness. And of course with the sun exposure, shelter and shade, Zurich only, let's see, availability of sunshine. We all note that it gets cold when it's the opposite here. Comparing cities, sound or sight of water. I guess if you choose to live in Zurich, you really appreciate the lake. And we saw from the results of places of vitality a grassy lawn. So maybe why Kina Visel is really popular. And thank you very much, but I guess let me go back to this for you to just take a look at a little bit. Yeah, so thank you.

All right. Hi everyone. I'm Matias Quintana and I'm here with my colleague, Dr. Javier Sanchez Baqueriso. I'm proud to present also, I think one of the most diverse modules all the way from Europe to South America. So we're happy to just advocate how FCL and SEC is very diverse in those terms as well. And without further ado, we'll introduce our amazing team. Some of them are in this room as well, some of them presented already, but this would have not been possible. The slides I will show you, if not by all researchers here in the second row. Some of them are abroad, but you can find many of them also in this room. And you will see their amazing work translated in outcomes such as papers and Other things, very briefly, our team, as also Christophe mentioned before, really gathers a lot of diverse expertise. We have data scientists, urban analytics people, neuroscientists as well. So we all come together to try to put some sense into these cities, whatever they are, and through the understanding and trying to map those in multiple ways and meanings that we can get out of them. So as we just saw in these modules today, yesterday, early in the morning as well, there's a common goal that we share of trying to understand cities and map cities. And we can see that there are different tools that exist through urban design, urban science, but sometimes they just stay on their own field or domain and there's really a lack of trying to intertwine them, to properly converge them into a single paradigm. This is where we want to put our emphasis and we put forward a concept that we'd like to share with you all. So in order to combine this urban science and urban design into a single, let's say, paradigm or multiple paradigms, we think there has to be at least a core of needs and requirements that these paradigms, lack of a better word, should have. First of all, of course, they should be based on reality of all actually, how cities are run, how cities operate, and we should be able to model those and try to get some meaning behind them. But these methods, these models have to be able to be validated with some, again, real data, with some empirical data as well. But all in all, these methods and its validations should be able to be spread out to allow for cross pollination of idea and to be communicated and transmitted among other researchers, among other dimensions. And therefore this transitivity requirement should be really a priority of this. One of these potential urban design and science approaches is what we would like to call a semantic urban element. There might be many, but we decided to come up with this one that in its core basically relies on an existing urban element that could be just an urban solution. Think of it as a pocket park, as a smart intersection here, but there's an added semantics to it. There's a knowledge model that is able to enhance what it already has as a tangible built environment with these urban design methodologies and insights that are only possible through the analysis of the kind of data it can have, how this data feeds into systems.

Such as data streams or whatnot, then power decision systems that could take actions by decision maker or will lead to better implementations, or whatever solution this smart intersection could have. Now, as we could talk a lot about the semantic urban elements on itself, we would like to show you more Precisely how this can be seen in different domains and particularly in three domains that we have explored in the last two years, primarily about urban analytics, urban design evaluation and knowledge representation, and finally computational social science, which Javier will explain a little bit later on. Now, in Urban analytics, let's say that as we mentioned, one thing of understanding the urban elements, such as that pocket part or intersection, we'll need to measure it somehow. One way would be through street view imagery and there are lots of it. Google, Street View, Mapillary. But what about that semantic data that we're lacking? There are sources out there that. But there was always, until now there was a need of properly putting them together into a single source to use both and get insights from both sides. The actual urban element that's seen represented by the figure and all the semantic knowledge, semantic data about it. This work led by Eugene Ho, who she is in the room right now, was an amazing effort spanning almost year and a half, if not two years of collecting data from 688 cities. It's a very auspicious number by the way, tallying up to 10 million images. And all of them have more than 300 metadata features each of them. This empower a lot of urban science research that could further into, for example, human centric applications such as urban perception and other type of analysis. But when it comes to urban perception, there was also a lack of standardizations that we still are trying to address with the concept of semantic urban elements. This standardization for urban perception was more around how do we even conduct the surveys? What questions do we ask? In which order? Which image should be shown? Just a panoramic image, static one in one direction and what not. Our work, led by our amazing intern Joe Long, goo also in this room as well, was thoroughly investigating what is the best way to conduct this service in order to get the best outputs that we need to further investigate urban visual perception as well as other human centric tasks. Both works already published, Feel free to use them in your future work as well as our amazing data set. Or you can ask more questions to Eugene later on. Now, building on these massive accomplishments of a lot of data collected, best practices to collect actual human surveys based on these images, we pose the question of does really perception and this human centricness of things vary across different dimensions, such as demographics, the place you're in, or even other features such as personality traits. It might be obvious to say, yeah, I come from Lima, Javier comes from Madrid. The same streets that we look at will be vastly different. I can see it as oh, I'm definitely going to get maxed over there. Javier is like, that's just Madrid. But this very inherent perception, biased on our upbringings, it is translated and how we try to analyze cities through the eyes of its pedestrians, through its citizens. So that's what we decided to measure with this and challenge existing studies that purely rely on a big, large, let's say machine learning model that takes an image and just infer a potential perception prediction without really accounting for that human nature of things. Here we demonstrate such things do exist, and it's not as easy as just taking this large model and infer that in your own city or multiple other cities. And why we do need to still keep collecting data from that particular location we care about for proper targeted interventions. Now, at the same time, we cannot forget our designers. At the end of it all, we are human design, urban design, urban science. And a lot of our efforts may have looked just to be purely in the science side, but. But all of these things can be easily translated to researchers in the urban practice. This work led again by our amazing postdoc Chenji Tsai, which is currently in a conference in Japan, that's why she couldn't attend. Dives into where is the best way then to encode all this information coming in multiple formats in an easier way, a more efficient way for practicers and designers to evaluate what cases would it matter for them based on their own queries and really look at the different types of data to make better informed decisions and really just leveraging that AI, that data side of things, other than expect them to code the whole thing in Python of something of sorts.

Thank you, Matthias. So we were talking at the beginning, Matthias was introducing the part of the analytics of full run analytics, the part of how we bring this type of insights for designers, for practitioners. And then we are talking now here about computational social science. And what is this about why we are bringing this type of interdisciplinary approach computational social science? And this is basically because we are caring about adaptable infrastructure. In this case, because we are bringing together not only technical scientific insights that we are learning from analytics, but as well we are trying to engage and to make it operational insights and knowledge for practitioners that we can bring for real in the city, but as well we need to make sense for policy making. And it's not only enough with technical insights with a good operation on how to implement that, but we need to bring together other type of considerations which are regarding human behavior, how people endorse, how people may react to this type of environments and Ethics, politics and all these other type of considerations that we have in urban settings. So this is basically like a very interdisciplinary, cross disciplinary and diverse type of scientific and practitioner based scientific endeavor. And that's why we are combining here many different methods. We are really on to mixed methods. So we are having like simulations, but as well we're having.

Experimental design and other type of approaches that allows us to engage with all these different types of knowledge for being able at the end of the day to be able to design, to envision, to propose alternative features for better cities.

And at the end, and this is one of the main focuses of the last part of the research, what we realize and what we are advocating for in the city is that one of the main issues is about access. Who has access, who has the right to access of the city and how we are distributing the type of access in a very limited space. Not only here in Singapore, that there are always, or you are always saying that it's like a very limited country, size wise, but it's a crucial issue in every single urban environment across the world. And we believe seriously that the city should be respond and should be able to adapt to these very changing demands and be able to allocate a fair use of space for everybody, depending on these needs. So we basically took to mine run two main experiments for being able to test these type of approaches about the insights of how the streetscapes were used in cities, for providing like a fair use, like a more diverse use. Because as well this aspect of accessibility, how we distribute the access or the right of ways, if you are coming much more from the transportation engineering point of view to the city is not only like a simplistic or like an innocent decision, but actually as well, you are determining what type of use, what type of activities you're going to be allowing for people. So it's like very important, very fundamental social and political decision that we as practitioners, as designers, as policymakers, we are making for the city. So we were posing here that probably we can try to accommodate. We are trying to design streetscapes such a way that we can allocate the space dynamically to different type of users. Because we know that the pace that the dynamics happen in the city, they are not permanent, they are not constant. But funny enough, we design the uranium space on a very static way. So it's not really matching to these different paces, to those different dynamics and rhythms that the city shows. So basically we develop this algorithm for being able, which is multi agent reinforcement learning for those who are like Much more machine learning for trying to see how we were able to based on these many different interactions that they are happening between different users of the streets, pedestrians, vehicles, cyclists, how we can consider their perception, their level of comfort, comfortability, level of service, of quality of the use of that space. And if we were able to show that allocating dynamically space on the streets, we're able to bring together a better streetscape for these users. So basically we were for sure we were having like efficiency goals. We are not mad and we know that people need to get around and people need to go from A to B. So we are not advocating for just saturating the city, but as well. And this is as well why we were saying that we are not only like engineers in this point of view, but we have much more holistic approach towards the design problem was how is safe, how is this comfortable for these different type of users? So can we dynamically share those streets, those public spaces for different users? And we developed this multi agent deep deterministic policy grading method. There are like on the table as well. You have a much more particular description of how this was designed. And basically like the adaptive algorithm show that we were able to outperform the level of service and the static allocation of the streets. So we were able to see that how we were able to dynamically allocate this space into the streets for all these different goals that we were envisioning. And interesting enough, we saw how co sharing and it was not like an objective per se. So it was not encoded per se in the algorithm. But how actually in some scenarios code sharing. So having like a much more diverse use of the street, meaning like you don't need to have like a huge hard segregation of different activities actually was successful on how to allocate these spaces. So how somehow having this decentralized is different considering each of those elements with all their interactions and all their meanings semantics, as Matthias was referring before we were able to improve this type of level of service. We can posit as well this type of question from another perspective as well, regarding from a much more technological point of view, with this type of new vehicles and agents and technologies for transportation coming into cities, which is do we really need to. Do we really need to believe or to trust on all these like massive adoption of autonomous vehicles and is that really needed or what we are envisioning even is it realistic to think about that nowadays? So following a similar approach and taking into account like a section of. I mean in this case we were working with a regular Grid of Barcelona because we had the data, so it was very convenient. But the model itself is very agnostic. So technically we can apply that to Singapore, to Zurich, to anywhere. We were trying to see how actually having a much smarter approach towards.

Transportation engineering, without relying that much on technology, we were able to have much of the expected benefits of all these trendy, exciting technologies about autonomous vehicles. Level five, level four, this type of technologies. So we create a number of policies for this very reduced amount of vehicles and for this infrastructure for see, can we do much more.

With less technology somehow? So how this multi agent algorithm can we able to provide the benefits without having like a huge adoption of these technologies of these autonomous vehicles technologies modeling different types of policies. So for example, how we were able to allocate dynamically this type of lanes for different type of vehicles. But as well optimizing the use of the line before intersections that as well I seen here that in Singapore wouldn't be like a bad thing to happen. And as well preserving at the end some of this resurface space for these smart technologies. As a result, we can obtain much of these benefits for smart technologies for transportation. Again like level 5, level 4 autonomous vehicles with a very small penetration rate of these vehicles according to these models. So, just to recap what we develop.

On those experiments by being able to learn from the insights from urban analytics.

I'm from the part of the how to operationalize on the work package. Second, how to operationalize that for designers we model on a much more granular level and taking into account policy, behavioral and ethics consideration about the allocation of public space, how we were able to show how these approaches were able to provide a much more effective use of the limited space of streets. And this comes together basically because just to understand the whole overarching idea that we were.

Pursuing in semantic Uran elements, on one hand we were talking about. Matias was talking about Uran analytics. Basically because we need better insights and richer insights for a richer modeling. In particularly when we are talking all the day around digital twins, we need a better representation about how we are able to represent all these many dynamics that are happening in two cities. Very complex interactions. Probably the most complex human creation is what cities are. Then on the urban design evaluation and knowledge representation we want to bring. We don't want to stop on the insights part, but we want to bring that in an usable operational way for designers, for practitioners, for architects. And finally on computational social science, we are positing that implementing these type of approaches are basically fundamental for being able to enable the exploration of alternative urban futures and being able to engage with the endless endeavor of making cities because it's never a finished product. Just to conclude.

We are saying for sure because we are using many of these models and simulations, that these tools are super powerful. But it doesn't mean that we are believing them as oracles that they are. We are not caring about their predictive power. But it's much more interesting because they allow us to think or to rethink and reposit questions about how things could be to think about alternative futures. And this is not actually what designers, practitioners, even policymakers, even politicians, I would say should be thinking of computation now. And this is different than what we were doing like 100 years ago, even 50 years ago, enables now engage with the complex interactions that in cities are happening.

Maybe we're falling short on how we are designing planning cities because we are trying to oversimplify them because of legacy practices in urban planning and design. And for sure complexity is not contained in a visual form on a render. And I'm not saying this type of like now super trendy generative AI tools are useful and powerful. I love them. I'm the first one who are using this type of generative AI tools for image creation. But the complexity which is underlying in cities is much more beyond this. And now we can operate with these complex interactions in cities, or we can try to approach them and instead of trying to have restrictive control or restrictive approaches towards how we plan cities and trying to limit that kind of interactions and trying to impose behavior on people and on users strategies, with this competition, we can use complexity in our own benefit.

Which can enable basically a series of basic principles. We can engage with very plural, very diverse type of people and agents and visions about the city. We can allow people really to express what they may need for and we can manage all this diversity in a meaningful way. And as well, as I was saying at the beginning, we can endorse this open ended nature of cities doing more. We can get much more by sometimes proposing much less and much more minimalistic interventions and being able to transition from purely social technical systems and even able to develop new urban theories. Because most of our urban theories right now that we are using the outcome from the 60s, they are coming from the 50s. We are not able really to develop new urban theories and we need them being able for being able to engage with well being, with health, with climate change and all these challenges that they are facing right now, basically for better future cities. Thank you so much.

Thank you Javier. Thank you Matthias, I now invite one speaker or one representative of each module to come to the front for the panel discussion. So with that, the second half of this session will start. And as you know, the overarching team is policy and governance. And we will come very soon to that. I just wanted to start off with one remark that I think cut across all themes besides the modeling. And so I think a lot of you mentioned that bringing together different domains and heard a lot, psychology, cognition, behavior, planning, design, just to name a few, is beneficial. I wanted to hear from you.

Why or in what sense. And maybe Ravel, you can start.

Okay.

I'm going to give in this case a kind of very personal opinion. I'm coming from architecture and urban planning. I was. So let's say this is recorded, right? Okay, that's fine. I was kind of dissatisfied on how from architecture and traditional practice and I love it. We're tackling many of the challenges or issues that we were having when designing a building, when redesigning a square. That for me was very surprising that why we are not using different type of methods. And for sure we are pressed by time, by budget, by many things in professional practice. But I was always kind of hungry of being able to have more tools or different insights from different areas for being able to inform in a much more effective way design. And that's why I think that we were coming together with all these very diverse mixed methods approach. It doesn't matter not being very rigid disciplinary about how you should tackle a project, but you can learn something from perception, from cognition, from psychology. Just endorse it, use it and reappropriate it for your own purposes.

Thank you.

Christoph. You combine cognitive science methods and architectural design.

Yeah. So again, I will work with a very personal answer. I start out with an observation. So maybe the three presentations today look like the three projects are doing very different things because we made different emphasis on more technology. More technology and then very qualitative data coming from survey and more behavior oriented data. Data maybe in the presentation that we gave initially. But it's basically they are all looking at very similar phenomena with a different mix of methods and each of the modules. And also the cluster only works with the mix of methods. And cognitive science, my home discipline is interdisciplinary in nature. It bridges everything from philosophy to computer science with psychology somewhere in the middle. And in a sense this is also how I perceive the projects in our cluster and we're trying to link from foundational research to application to influencing the future. And you don't understand the nature of human Behavior, human needs, human cognition with just one approach. I'm a psychologist by original training, and psychology is not enough.

Right.

And also it's a very long journey. And this is my favorite prop. I always bring it everywhere. It is about a conference in architectural psychology. That's what it was called back in the day that was held three months before we landed on the moon. Right. That was before my birth. And all of the important questions are in there. But in the late 60s, you were critical about the 50s and 60s. In the late 60s, the questions were there, but the ways of answering the questions were not there. The kinds of data that you see presented by these three modules.

Is very computational, but it is also deeply rooted in talking to people, looking at people, interacting with people. And this is how this comes together.

Felicity, you have these different work packages on urban data interaction efficiency and comparative urban vitality.

What is the benefit? What are the challenges and how you bring both together to make sense?

Well, I think this is, again, I also say a very personal story, basically.

Just reflecting on what has been in the last two and a half years for our project. Initially, when Markus Schleffer approached me about this project, he was the original PI. He said, we are doing this and we are looking for somebody in urban planning. So I said, okay. I mean, I'm trained in urban planning and urban planning has always been interdisciplinary. So I said, okay, this is super attractive. I get to meet computer scientists, physicists, and then I bring in my urban planning, and then we can have a really good conversation and really understand how some of these things come together. Well, the reality is that the practical.

Scenario of those conditions are working out, that integration is not so straightforward. While I think we all got insights into how the scientists who are doing physics and computer science think, and they have in some ways also be informed by our perspectives that look at more qualitative view of the city. I think working together has its own challenges, and so part of it is also time, because I think when we started the project, it was obviously quite ambitious to put things together.

But.

In some ways the qualitative part that.

Was thought to be part of this program would take the results for the quantitative and then look at it. But things in life don't happen sequentially. So I think this is one of the challenge. But I think we have all gained new perspectives and insight.

What I hear from all of your three very personal testimonies is that breaking up silos is actually the way how we can produce new and advance new knowledge.

Which is a very interesting point. And I also totally agree. I was also struck by all of you used the word evidence based research. And maybe you can say something on that because this also leads a bit to what I actually want to talk about to how can we make or translate research into policy or make it applicable in different ways and maybe first something on evidence based research in architecture.

I would say that is.

Kind of very common word actually considering that we are not really quantitative people. I have many discussions with Matthias about the nature of architects. Evidence based is.

A part of the practice base, but as well as to show and to be like as well. Very pragmatic and very practical and apply and being close to the actual issues or close to the actual challenges that we may be finding in architecture. So it's not for disregarding other type of research, but the research that we are doing, particularly at fcl. I think that here I'm talking for most of the projects that I know and the modules is really attached and grounded into actual issues. And in that sense, like evidence based mean in our context or in particular in the context of su, is that we are really working with cities, we are working really with.

Real data, with real locations, and funnily enough, we are discovering things which are just out there and they are able to inform part of the vision of part of the assessment or evaluation that we are doing for being able to tackle design, for tackle planning, for informing policy. So in that sense it's like very grounded. It's not we are just therapizing about abstract aspects, about cities that they are nowhere, but actually they are very grounded, very tangible elements that they are out there. And in our case, evidence based is really connected to a very grounded type of research that can be useful for Singapore, for Zurich, but for many other locations.

Can I pick this up?

I think the term evidence based is a loaded term, first of all, and it has a different.

Power in maybe more urban questions, on planning questions than it has in architectural design. And I think in architecture this logic of evidence has always been difficult because there is a lot of evidence that is different in architecture, that is different from how scientists are used to operate. There's a lot of this tacit practical knowledge. There's a deep understanding of the human condition in any good architect without necessarily needing the science. And very often when evidence based behavior science comes into the architectural domain, the request at the other end from the practitioners is more for a general philosophical inspiration. Think biophilic design, those kind of things. It's more design principles than the kind of evidence that maybe in Other domains of designing technical systems like human computer interaction, you would get. And one of the aims that we in our research, when we come to the translational have is to, I mean, again, going back to the terminology of the 60s and 70s, there was all this post occupancy evaluation, right? Which always comes late as the term implies. And one of the aims that we have is step by step in our translational efforts to come closer to pre occupancy evaluation, to evaluate not design principles and philosophies only, but in a concrete setting. And this is for example, what Yujin did with UN Studio. They designed different options for these transport hubs in Austin and then in virtual reality, tested how people respond to those. The virtual reality simulation is not the same as the reality on a hot summer day in downtown Austin, Texas. But if you compare different designs, you can look for visual appeal. What attracts movement? Where would people gather? And you do this in the concrete case of a design that is implemented, that will be implemented where the stakeholders have a buy in, and in this case, where the stakeholders paid actual money to get this research done.

Can I just quickly follow up on both of your points? In the sense that I think before we touched upon this in the previous panel, but not really so deep, is like there's still a difference between the simulations or scenarios that you model and there's still a big gap from that to implementation, like in a real world or in like, let's say in a governance setting, which is much more like tied to processes and steps in the bureaucracy. Right, like, so.

There'S still a lot of translation that would be required, which is beyond the whole modeling and scenarios, right? And I think that's what also Einar showed. You showed different options and you showed the surveys, but then still this knowledge needs to be picked up, right?

Well, I'm going to say that a good question because I think what you're putting your finger on is that tension between science and policy, right? And you know that should science be influenced by policy solely? And should policy be made fully just based on science? So I think this is my personal perspective. I think in urban planning, which I'm trained in, most of our questions are very aspirational, right? And practical. So we want to make livable cities, sustainable cities, resilient cities. Now the question is, how do we go about studying them? And so I think what we've done in our work package is that we take that and we move it into the scientific question and to ask, I mean, and from there a practical question. And I Think this is something that I do struggle as researcher but also trained as an urban planner and have practiced before that oftentimes the findings the planners will be like what can I do with them? So there's still a need to translate that. So that's why we ask the question how can we enable planners and designers to take some of these. So I think it's that thing perhaps that connection is to keep in mind what is it that is needed by the policy and planners, the policymakers and planners. And how can I ask my question, the scientific questions in a way to then do that bridging. But it's not perfect. And we always try to make sure that we are doing science that is not completely just policy and policy that cannot completely just be science per se. So yeah, so it's a really tough question and I think we all face this practical constraints and challenge. Yeah, thank you.

Someone else wants to answer.

I'm happy to follow up to that. I think it's largely is also a matter of the spatial scale at which we're talking. Right. So it's these policy questions are much more clear the moment we step out of the building. The moment we become especially especially in Singapore when we talk about districts, when we talk about estates, when we talk about the space between buildings there. We also collaborate with other modules here on Cities of Tomorrow, grand calls, et cetera, to look at these spaces that are strongly mandated by different government agencies. The moment you step inside the building, you step into architecture space, you'd step into very mature discipline where almost all things that are vital to pick up the term of vitality earlier today are already mandated. Right. We know exactly what the fire escape code is, but there is no mandated measure of how much interactions does an office building have to offer. But increasingly we see in brand name architecture companies across the world and that they similar to how they maybe some of them did it in the 70s, 80s and 90s are having a new wave of bringing behavior scientists into their teams. And we're working with these people in industry, Ewan Studio, a couple of others Foster, they all do this. They bring behavior scientists in not in order to come up with better code. They try to find more human oriented solutions and different ways of inspiring their design choices.

That would be my other point, like going into human centered design. But I see we have half an hour left and I want first to open up the discussion before I monopolize here the round. So if there's questions from the audience.

We could now move over.

If not, I'm happy To continue.

Human centered design.

In your fields, we often hear agents, users.

Terms or categories that classify humans in a specific way.

What do you think? And I think you hinted to, I think Stravko, that we want to. Or the next step is to put the human in the center. Maybe you want to elaborate on that or Christoph.

Well, I can pick it up first and then Stravko can perhaps given a microphone, add to it. Again, I think the term human centric is used in many different ways. Right. And I would really like. I'm not an architect, but I would really like to emphasize anything an architect does is also human centric. It's not like they need psychologists to come along to finally think about humans. But at the same time, it's of course true that architecture is a multi constraint planning problem. So you have other things to think of as well. The side, the money, structural integrity, the not so human centric desires of your clients sometimes, right. Prestige, all of that comes together. And I think by constantly at this interface between science and society and the design professions, by constantly emphasizing this human centric, we maybe try to sometimes tilt the balance of priorities a tiny bit more to the people who actually work, live, engage in the buildings rather than those who pay for it or design it. That would be the first part of it. But Stravko, please, as an architect, no.

Obviously you summarized it well, but the idea here was really not about.

Saying that the human centered design doesn't include, I mean, see humans in the right way. It's more about.

Kind of critically thinking or repositioning ourselves within the module and where in that triangle that we showed at the very first, where the people come in, once we start dealing with architectural and cognitive science and to what extent the humans in our research or whatever we did were objects of study or someone to design for and things like that, where they were still in a way a bit passive if you have to be critical about what we do. So the next step would be really to.

More actively engage them in all of these processes. And it's also not something that is completely new. There are many of such attempts, you know, participatory design and so forth. Is this working? Yes.

So it's not like we are, you know, reinventing the warm water. Right. But it's really about making whatever we do as inclusive and as intuitive as possible while still working with evidence. I mean, that is a tricky part. And also cutting across different scales from how I see, how I touch, how I feel, to the planning issues of where do I put A train station and so forth. I mean, these are really difficult things to kind of transgress. And this is where we sometimes forget the scale, forget that human person where we call agent or user. Suddenly it's anonymized, but at the end of the day it's not, not. So this is where we try to kind of put a little bit of accent.

I can follow up a bit and I want to connect with both of your last questions regarding the part of the policy.

Political, societal challenges of bringing planning decisions, but as well with the human center. Because it's very related to the computational social science approach. Like with this much more holistic and complex.

Interactions that actually you are facing in two cities. Not only about the idealistic.

Close environment of just the design practice, but at the end you need to face and you need to interface with many other instances, both when you are considering users. I don't like to use the word users when we are talking about people who are living in the city. It's a political statement.

But a part of that. And that's why as well, we were thinking, and I think that is very related with the work of urban analytics. We need to stop to think like a people, like generic, without shape bodies of people. But actually now we have the tools, we have the technology, we have the capacity really to understand or to try to differentiate individual behaviors, individual preferences. That's as well human centric. And this part of what we were trying to start to do in part of our work as well, we are trying to understand what from the professional practice and as well as people. It's like human centric, like what designers or what experts are doing. But then the other challenge that I was mentioning in the presentation was precisely okay, we can have this idealistic, very aspirational way of facing urban design challenges or building design. But then we need to be realistic that we have developers, we have political interests, we have many other interfaces going on in the city. And we need as well to understand, to model, to be enabled how we need to intervene and to interface with all these other instances in the city we need. Which as well is like another perspective of human centric research. Maybe you are not focusing on the inhabitants of the city, but you are focusing in this case with people who have power for making decisions or who has like any other agendas that even they are contradictory or opposing to your own agendas. But you need to be able how to play and how to deal and how to work with these conditions.

So do I understand you right, that one of the things which is our job or how you described it, is to make urban knowledge more accessible to.

Yeah, I mean definitely shouldn't be like an enclosed like expert knowledge, like a black box for people who are not experts. It should be able to be transparent for citizens and not patronizing, which is as well like what we observe or we have been observing a lot in participatory approaches that there are like a lot of patronizing behavior attitude by experts and we should avoid that. But as well for politicians or for other city officers that as well they may have like very insightful local knowledge or being aware of other conditions that maybe you are not aware as a designer.

I look once more into the audience. I know I see many hands.

I saw three there and one there. So maybe the first. Yeah.

Okay. Camilla from ETH Zurich. I would like to touch upon something that you were just talking about again because you. I feel like you're maybe underselling your own research. Research a little bit.

Because the question of whether like human centric design is something different. You said like, no, it's not that other types of design don't consider humans. But I would like to argue with this a little bit and I would like to ask all of you in the panel about your opinion on this because I feel like as a society we might value some things more than others. And like, for example, just to give you an example, the building in which I'm sitting in, in th Zurich is a very new building. Nobody really likes it.

Nobody really likes sitting in it. I mean, because from the outside it looks amazing. Right.

Which building are you in? Just to contextualize your question.

Hif.

I'm sorry?

Hif. It's like super modern, super nice. It looks like a Dexter's Laboratory basically. But it's not really nice to sit in there. And so basically my question is related to values because I feel like this building is a really nice building to have on display. Because if we think of our value as like we want to show our institution as, you know, forward looking with this nice design and so on, if this is what we value, then this is the right kind of buildings to build. But if we want something to be comfortable for our researchers to work in, it's a little bit different. Right. And the same thing goes when it comes to like the value of technologies, of new technologies. Because there is a lot of research on buildings that's very techno oriented, very techno optimistic and humans, I feel are sometimes lost a little bit in there. So there is not as much focus on societal values on Community. We talked about it a little bit yesterday in some of the sessions, but it's still somewhat on the outskirts of what we're talking about. So, yeah, I would like to ask you about this very general question. Sorry if it's a little bit too general, but about the values that we have when we design things and why don't we do more stuff for humans. Thanks.

Well, then I'll pick it up and maybe Felicity wants to add to that. So, first of all, underselling. I think in interdisciplinary research it's really important not to overstate. Also trying to become a Swiss person now. I also try to not overstate.

I'm originally German and Germans tend to maybe not be as humble.

But taking your example seriously, I'm not entirely sure if I map HIF to the right building now, but there's numerous of such buildings. Right?

Say again? Oh, yeah, that one. Yeah, yeah, yeah. I've been in it and it has many great things about it. And I've also been in some of the meeting rooms and they were not as practical as they could have been. Right. And you will also find examples on other campuses of other universities that are like this. But this is what I meant when I say multi constraint planning problem. So there's many decision makers and many dimensions that come in and I think also Felicity's work that's more qualitatively driven and also work that Drafko is driving in our team, where we also are increasingly interested really in this, in almost a phenomenological perspective. So if you go in into a setting like this and. And look closely and talk to the people and sit there for a while, we always argue about putting yourselves into the shoes of the other person.

Then you become quite sensitive to all of these elements. But having said that, if Etihad were to build a building that is super cozy, that kind of would be off brand. So there's a certain tension there.

And yes, so I think as a program and as individual modules, we always navigate such tensions.

But I'm more optimistic than I was a few years ago because I increasingly engage with practitioners in industry and they want to make money and they have to convince their customers. Right. And the customers come. And just to share another example from our collaboration with UnStudio, one of their more prominent projects in the last few years was the headquarters for booking.com. right. And they put a lot of emphasis on now also post occupancy evaluation, finding out what interactions work well, what interactions work less well in this building. And there's There is a curiosity of the client and a curiosity of the design company to finding out whether the aspirations are met. And yes, it's more difficult in architectural design than maybe in digital design to change something after the fact, but there are ways of dealing with this. And already this brings us to the topic of appropriation, which is one of the keywords for our hopefully our work in future cities. Lab four right. Look, how, how do people make use of the spaces also in ways that maybe were not intended? Right. And, and if a meeting room doesn't function, or if the balance between privacy and, and communication in an office building doesn't work, then people come up with certain types of solutions, right? They put in an extra blind, they turn the desk differently and, and they improvise. And that again, speaking as a scientist, again, is the kind of trace of human needs that you can pick up. And an important difference just to finish off this idea is that very often we look at stated preferences, we ask people what they want. But if I come to your workplace and I see that you have turned your desk differently and that you've put on this extra blind, then I get your revealed preference, then I can see and instead what is really hurting you. And that I think is extremely valuable. And I, and I see an interest in picking this up in a lot of the modules in our program. And you can use quantitative and more qualitative ways of getting there.

I agree.

No, I mean I wanted to follow up, but in the sense that like qualitative and quantitative and the mixed methods that have been shown, especially in this I think cluster, which are very strong, they also are informed by different data. And we could also go into that one, but before doing so, and I don't want, if you have like answers ready, I want to hear like more questions and maybe like they combine with what has been said. I saw two and there as well. Yeah, please.

Hello, can you hear me? Yes, hello. Thanks for your presentations. I'm Luciano from sutd. So I'm curious about how you think your research could engage more with the private sector, essentially from other sources of funding beyond agencies, and especially focusing on small and medium architectural fund firms, which in most of the cases don't have neither the funding to pay for research as service nor the time to do that by themselves, despite being a possibility in the form of a ballot proposition, as in the case that was pointed out of UN Studio, which is a very large scale firm.

But however small and medium scale firms do need that kind of information in a very fine grain delivery and very specific information essentially for morphogenesis or at least for layout development or strategic design. So do you see the trajectory of your work as researchers catering to that audience? And if so, how would you proceed? Thanks.

Thank you for your question. It okay, indeed. Like a short answer. Yes, many of the small scale firms.

And not necessarily only referring to necessarily boutique architecture. So other type of small firms that they are doing like probably like 90% of the buildings that we are seeing out there, they need that. The things like here we have like a few things. First of all, I am aware of tech firms or kind of startups that actually offer this type of service or capture this type of service for smaller firms that for sure they are not able to have in house knowledge. So I'm aware about a few companies like in LIS right now that I can think like in Australia and in Norway and Sweden if I'm not wrong right now I don't know if all of them actually they are doing well. I don't know they're accounting but they were existing at least two, three years ago.

And they are finding kind of like a business model for being able to provide these services. And I'm assuming that they are able to provide value to those firms and those firms are able to communicate and to translate that value to the clients. And now it's a different question here, like which is the value of the architecture practice? And that would be like a completely type of research question which is the value of architects firms and urban planner firms working for developers for. Working for institutions for who is paying basically the party. Let's say that. And that's another open question because at the end in many of the cases you see that you have so many constraints regarding building code budget, just selling your product to the, to the final user, whoever is going to purchase your office space that you don't have. Or it's very complicated, very challenging to those firms to be able to really find like an opportunity space. But this is the thing like many of these tools or many of these knowledge is not particularly complicated. It's not particularly hard to implement into design. What happens in many cases is actually we just don't know, we don't have the data or we are not able to provide these key elements that could.

Be.

Add to the value proposition of that firm. So in that sense, yeah there is like that would be a way of actually communicating and translating this research or these insights.

To that community as well to practitioners who are maybe working. These are the things like many of these people who are working on these companies. They are architects themselves. But instead of actually capturing or working on their own design firm designing buildings, they just decide that hey, this is like a much more interesting field to explore on this type of technological IT services for, for other architects. I'm still doing design, we can communicate to that. And here there is like another.

Challenge that we have about how we communicate our results, how we communicate our research. And actually it was a conversation with Jacob from Munster. There's like architects for sure and many of the designers, they are not reading papers in many cases. I mean, it's delusional that we are expecting that we're going to have like a world class research and they're going to be reading papers. It doesn't mean that the papers are not important, they are super important too. But as well we need to tune or we've been able to translate those papers and those insights in another type of communication that is able to be transferred as well to these other parts of the architectural practice. And that's as well probably like one thing that should be further explored and further tune in our, in our activity.

Sure.

Maybe directly pick up from this side. So this moves a bit away from the questions of how do we do research to how do we do all the other things that we are trying to deliver as individuals, as modules and as a program. Right. So this touches upon outreach, it touches upon teaching, it touches on impact. And so we do a number of different things that yes, we have done this only with a select number of companies so far, but we offer placements, we invite people to come to us for various lengths of stay. We offer ways of publishing together. Right. Which would also make cost sharing more viable. We place our students in industry. We as a collective, as a program, have developed teaching for agencies. One could think of doing that more practitioner oriented for people in industry. If you were asking me this question in Zurich, I would emphasize that we will have all these certificate of advanced studies programs that would be a way of going. So I think there's a lot of opportunity. And one of the things that we always volunteer if somebody comes with a nice case, is that we come out to them for a week or two and we camp at their side and do a study together. Right. And this is a privilege that we have as scientists that we can go out and respond when there is an interest. But it's also true, expecting non academics to read journal papers, even the ones with the highest impact factor, that's delusional. We have to find different ways. We need to do These brochures, we need to have websites, we need, need to have all of these other modes of communication to engage and to activate. Yeah.

And in a way, it also mirrors the complexity of stakeholders that are present in urban affairs, broadly speaking. And it also shows that the pipeline between science and policy is not a linear one or even the nexus between science and society.

There were more questions. I see two hands and then you. Next.

Hello. Yep. So this is kind of going back to the previous thread about remembering the person in human centric design and qualitative methods. So I was quite heartened to hear that. In terms of this new or more recent attention towards kind of putting aside the terms agent and user and remembering that it's really a person there. So it was great to hear that. And it seems, of course, that that was not always as much the case. I remember like two or three years ago, I was participating in a research study where they had EEG and eye tracking and they were studying how we reacted in urban versus green environments, urban versus green environments. And they never once asked, what are you actually thinking? And so I asked them.

Are you going to ask me anything? And they were like, oh, no. They just laughed. Why would we want to know that? And of course, I bring this up not as a slight or anything, but just to illustrate that there seems to have been a shift in how we think about research and people. And of course there's the tension between stated versus observed preferences, as mentioned there. So there's that. But in a broad sense, of course, even in terms of public participation and consultation, now there seems to be a new focus on talking to the person and qualitative research, of course, that seems to still be developing. So I'd like to get your thoughts on the opportunities, challenges of kind of incorporating qualitative research and talking to people going forward. Because of course, at present it seems, I heard there was a bit of a patronizing attitude that was mentioned sometimes in public consultation. And so going forth beyond just public consultation, how do we really get them involved in research or even in research design studies, having qualitative research methods where we talk or observe more in depth and engage with the person, not just as a research subject. Thank you.

Oh, well. So first of all, it's not impossible that the study you were in was linked to research from our own team. And if you felt patronized by the research, I am deeply sorry and I'm happy to have a conversation about that.

What we try to do is engage in a kind of triangulation between qualitative and Quantitative data, because purely qualitative data is also always very subjective in how you respond to it and how you interpret it at the same time. And this is really a critical challenge for anything behavior science, be it computational, social science or psychology or cognitive neuroscience for that matter. The things that we can numerically express are always proxies of the things we're really interested in. That's the one thing. And the other thing, and this is also something where qualitative research can make us more sensitive, is that because we translate all the phenomena in the world into numbers, then the statistics we use tend to focus on the central tendencies, the means, rather than the variance. Very often it's super informative to look at the variance. The fact that people are not all the same and that there are people who have special needs and there are people who have different attitudes. And we, in some of our research, these go a long way to activate other parts of society as well. It's so much more tedious to do or difficult in terms of conducting research to have as participants in your real world studies in the neighborhood some elderly ladies who may be a bit frail, but understanding how they behave in their neighborhood is vital if we want neighborhoods to cater to older people. So we have to broaden the participants and we have to broaden the ways we look at it. And also at Ethe Herzur, I'm sitting in the ethics committee, right? And there's a big point being made of the fact that the people who participate in our research, they co create the research, they are not just object of study, they are not our subjects, they are our interlocutors, they are participants and they are actively engaged in research. That does not always work out because we are all just humans and we all make mistakes. But the aspiration is key, clear to engage.

Sure, I can say something about qualitative research, since I do that. So for this particular work package that we have done, we adopted an online survey. And it's pretty big. It's over 1000 data points. This is much bigger than the kind of research I normally do, which is actually interviews, cognitive mapping, spending an hour plus with people. So maybe I can talk a little bit about the difference, right? I mean, from one perspective, this kind of online survey is already considered qualitative. But then the other spectrum is you actually spend one and a half hours for the people on the online survey. I do not know who they are. They are anonymous. And so we got a lot of bots, right? So one of them was cleaning the data. And I would Say after having done this, would I choose to do it again? I think there are many questions that I wish I have those people to ask and ask them, are you for real? That I don't get a chance to. I just have to imagine they are real. We cleaned it as much as we could. I want to give an example. For example, there was a data point that said that they are in their 50s or something like that, they live in the city center in Singapore, they live in the landed housing, they have a university degree and they earn less than $2,000. These are things that's like, is this possible? Should I still take their responses as valid? So these are real challenges that doing even something like an online survey, which is qualitative in some nature. I think what I have to say is that the other research that I do which looks at mapping urban belonging in places that are undergoing change and facing urban redevelopment, we do cognitive mapping there. And what I really do enjoy for that kind of depth of work is that there is accountability because you spend time with the person, you really hear what they say, you have questions you can so called triangulate from the different interview questions you have and eventually the social demographic data that you collect, you know that you've been talking to them for one hour. They are not going to, you know, try to fake their gender, fake different things. So.

But you know, that took one and a half hours. And it's just one data point and I think this is always a challenge in policy is that they are looking at numbers, right numbers to make decisions for as many people. And qualitative research has its limitation in that way. But nevertheless, I think the fact that you can actually get accountability in your data point, you spoke to real people. And I couldn't more emphasize that during an exchange when you do an interview, someone, something is changed, right? You know, that time you spend with person, you are changed, that person is changed in some ways of looking at the phenomena. And I think that from all the experience I've had doing interviews, most people say thank you for taking the time to listen to what I have to say to allow me to do the reflection. And even though I think maybe this is a good response is that we say this is the voucher for all your time. And you go like, no, no, no, please take it back. And of course we couldn't take it back because that wouldn't be right in terms of the financial accounting. But I think what that shows is that they've enjoyed the time. I think they appreciated that someone stopped and Talked to them. Of course, it's a bit of a self selection. Some people would never want to talk to you. But nevertheless, I think we've got insights from that project, from Belonging, 165 people in five neighborhoods in Singapore. And in this case I think we've got some kind of insights. Different scale, but nonetheless this challenge in doing quantitative work.

Thank you Felicity for your answer. Since we started later, we have still time for one last question in the third row.

Hello, I wonder the first tank, y'.

All for the presentations.

I really enjoyed it and I wanted to expand more on the human behaviors and the human, well, beings that goes into these like design solutions. So when actually considering like human beings or like people in like a certain environment, especially if they live there long enough, their behaviors are actually bound to change and actually preferences even might actually differ. So in regards to these kind of like changes in behaviors and preferences, what are some like design solutions that you are currently heading or thinking of?

Who wants to take it? Short answer?

I mean, were you addressing any of the particular projects, mostly like generic.

Yeah, it's a very broad question, so like anyone can answer.

I'm not even sure that I fully agree with your premise. Right. That behavior. Well, some behaviors change, some behaviors stay constant. And then the question is why would they. Right. So. So for example, in the study that you showed about how people deal with density and buzz and vitality and that. Right. Some of that clearly is of a shorter term nature because you only go to a certain amenity ever so often. Otherwise it maybe gets boring at some point, but then at different timescales it might become attractive again. And also, and this brings us back to this matter of not looking at the average person, but the variability of people. And some people want more change and some people are more static in their behaviors and it has to do again with values. So for some people this stability is a priority. While very often, especially when we come from a very urban background, very designy, we want to change. So it's really difficult to make that change choice not to do anything in terms of rebuilding, for example. So it requires talking to a lot of people and getting a sense of this variability, I would think.

Yes, I think that's a very good closing word. Thank you for your attention. Thank you for the presentations. I think it was very nice.