 Welcome, everyone. Let's talk about urban planning. Today's topic, Cape Towns Maze-like Informal Settlements, which have unique movement patterns. Experts, like our speaker today, Jalil Borowski, are rethinking how people move in cities, because in this case, the usual rules do not apply. And revisiting these rules could lead to better lives for residents by making sure that everyone can easily get around, easily feel connected. Jalil, welcome to our episode. Thank you so much for having me. Jalil, what made you study specifically Cape Towns Informal Settlements? So this paper comes out of actually a PhD research project that was actually very much focused on the challenge of public lighting and informal settlements. And it was a collaborative project with other colleagues at ETH Zurich, where I did my PhD and where I'm currently based. And we were actually very interested in the problem of the existing public lighting in informal settlements and how to solve that problem. And what comes out of an interest in public lighting is a question of, where are people at night when they need public lighting? And this is kind of the entryway into this interest in what are pedestrian activity movement patterns in these places. It's actually something that there's very little literature on. Exactly. Let's follow up on that. That's what I was going to ask after. So what was specifically missing in the research shot there before you published your article and prepared this study? So I would be happy to be wrong about this, but as far as we know, there's really no other study using sensors and or collecting quantitative data on pedestrian activity patterns and informal settlements, particularly at night. So there are some papers that qualitatively measure, well, qualitatively and quantitatively measure pedestrian activity movement patterns through manual counting. So they'll stand outside and count how many people pass during a certain window of time. And this is really useful information, but it's very general and it's really only collected during daytime from every paper I've seen. This means that about half of a 24-hour period we know next to nothing about. Yeah. Perfect. Well, let's then jump into your findings. Sure. So I guess maybe to just give a bit of context, what we are doing is we're empirically measuring pedestrian activity between 6 PM and 8 AM in the morning with sensors that are installed on people's homes. So when you pass the sensor, it records a trigger that someone has passed. And that's how we formulate our data. And we really wanted to understand two things. First, just what are the patterns of pedestrian movement at nighttime in this informal settlement? And secondly, how do existing theories of pedestrian activity compare to that data? So first, in terms of the basic patterns, we really learned that activity is highest in the early morning. And also in the evening, kind of when people are coming home from work, as well as on the weekends compared to the weekdays. And this tracks with what we actually know from our qualitative work and from survey-based research we've done in this same neighborhood, which is that many people rise quite early in order to go to work or they do a night shift and come back from work very early in the morning. And that regardless of work schedule, many people are more likely to be home in the evening and on the weekends. And we can see this in the data. When we compare our empirical data to these theory-based predictions of pedestrian motion, we have two that we refer to. First is the shortest path analysis from root optimization, which kind of says that people take the shortest path. And then there's a space syntax analysis of choice, which basically tries to calculate which paths are most used in a network. When we compare our empirical data to these theory-based predictions, what we find is that there is some correlation in path usage between shortest path predictions. It's weekly correlated with observed evening activity, but not at all with early morning activity. And it seems to perform equally well in terms of weak correlation on the weekends and weekdays. On the other hand, when we find no significant correlation at all between the space syntax measure of choice and our sensor measurement. So what this kind of means is that these very network-based theories of pedestrian activity appear to have kind of limited relevance in informal settlements. These aren't the only theories of pedestrian activity in urban planning, but they are the ones that are kind of most known for being based on the configuration of the network, which really is most comparable to the data that we're measuring. The sensors don't collect any other information about people besides that someone walked by. And so these are kind of the most comparable to our empirical measurements. Maybe just in addition, I would mention that one other notable finding, and this was kind of why we wanted to publish this paper in the particular issue of urban planning that we did, is that we really found that both theories did poorly at the entrances and exits to the informal settlement. So the informal settlement is surrounded by a formal area. And these kind of entry points or exit points is actually the area where the predictions and the data actually matched up the least. It's possible that this is a limitation of our work. It's difficult to completely rule that out. But what we actually think is that it could also be speaking to the fact that there's really a gap in theory for understanding how formal and informal connect to each other. Exactly. And on that gap, so some correlations found others not between your empirical case and the theory and, in this case, to informal settlements. So I assume that this revisiting of ideas might concern policymakers, city planners, individuals. So tell us more about that. Yeah, so I don't want to overinterpret the results of one case study. This was like, I want to reiterate that this was data collected in one informal settlement. But I do want to highlight a couple of thoughts, which is first that access to all kinds of infrastructure and informal settlements is pretty much across the board a serious problem, not to mention issues with disaster risk management, all of which require planning and policy. And these kinds of decisions can be heavily informed by where people are and which paths they use. And that currently seems to not be information that's very easy to get a hold of for planners. So I think this paper kind of shows a way to do that and that it can be very insightful. I mean, just to give the example of toilets, which is a very big issue in the informal settlements in Cape Town, in many informal settlements, access to toilets is through some sort of shared sanitation infrastructure, which means residents have to walk outside at night to use the toilet. And so you can see how understanding which paths people prefer to use would really maybe inform decisions about where to site these types of shared infrastructure to make it so that it's more easily accessible for the most homeless people. And I just also want to say that our study also highlight the importance of how studying these patterns could really also improve upgrading plans, so not just responses to the infrastructure needs and informal settlements, but also plans for upgrading, which is part of informal settlement policy and planning in some cities. Our study also shows that it's really important to be careful when applying theory that's predominantly tested in very different contexts to informal settlements. And this is really important for policymakers and practitioners because they're the ones who take the theory to the field. And I think that this can be really challenging because informal settlements have a set of characteristics that are very different from the formal areas of cities. And I think this study really highlights that. Another takeaway is that while sensors have several limitations, especially in informal settlements, and I'm happy to talk about that, they also have the potential to confer large policy and planning benefits that could really improve the way people experience informal settlements, particularly in settings where it's otherwise difficult to collect information, such as at night time. If you don't live in an informal settlement, it can be very challenging to work there at night. And these sensors are kind of a nice conduit for residents to be able to kind of make known what is going on, what is their experience at this time of day. So I think these are kind of the most key takeaways for urban planners working in the informal settlement space. And maybe just my last point on that is what we really learn from our broader research is that night time is a real serious source of worry and stress for the residents of the informal settlement we study. I suspect that's true for residents of other informal settlements. It really seriously impacts their perceived quality of life, possibly also their physical and mental health. And so I think this is a major call for policy to address this through better understanding of how people are moving in these spaces at night. No, that's great. Is this policy impacts that makes, well, and your reflections, has you did that makes these episodes very, very rich. I want to take a couple steps back to what you said before, and actually several times. And I want to follow up on that. She indicates in your study, because when I read it and hear in our conversation some issues and limits to sensor data, some time limitations, some study area size that require further investigation. So let us know more about research limits of your study and what's ahead of us to study. Yeah, so maybe I'll highlight especially this in the limitations linked to the sensors because I think those are a bit less dry. So like I mentioned before, our pedestrian motion sensors essentially work by detecting thermal differentials. So that means that they basically detect if someone or an animal could be two, but mostly people with body heat pass the sensor. It then senses, OK, up here's someone with a temperature above a certain level, and it causes a trigger. So despite several pilot tests in informal settlements, in this informal settlement, the final sensor that we ultimately implemented in the end could not collect accurate data during the day. And it was really difficult to figure out what exactly went wrong. But we believe that the reason is because of thermal interference from the surrounding building materials. And so basically after a lot of testing and validation, we realize we can only really be certain of the data in the early evening hours into the early morning hours. So between kind of civil twilight times. Therefore, we don't have empirical data for daytime. That means that even though we were interested in studying nighttime without daytime to compare to, we have kind of a little bit of a black spot, not a little bit. It means that we really can't say if the theories that we tested only fail, so to speak, at night, or if the weak or absent correlation actually holds over the whole 24 hour period. And I think that would be something I would really like to be able to fix about the study. In addition, while we intended to install enough sensors to cover the whole path network in this neighborhood, some sensors were stolen, some sensors didn't function properly, and therefore we have missing data for certain parts of the network. And while I suspect that this would not have a large influence, we can't say that for sure. And that's important to be really transparent about. We don't know if the correlations would somehow be different if we had a measurement for every single segment within the network. That's just the reality of testing out something new. There are some limitations also linked to how we calculated the theory-based predictions. I think it's a bit dry. So I would encourage viewers to read the paper. But I will say that one challenge is really defining your study boundary. So we didn't just limit it to the informal settlement when we did the theory-based predictions. However, when we measure with sensors, we only measured the path network within the boundary of the informal settlement. And this can have implications for how the calculations are done. Well, not how that we actually execute the method, but how the calculations turn out. It can, depending on the size, especially with space syntax, depending on the size of the study area, I think you can get quite different results. And I think this is something that we try to make very clear in the paper. And why? When I talk about these very interesting findings at the border, I think this is a good jumping off point for more research. Because I don't think it's very well theorized what is happening at the border between formal and informal areas, and how do these connections or lack of connection work. But it could also be that we would see different results if we had used a different kind of boundary box for the geography that we're looking at. So not only some tips for future research, but apparently many paths still I had to find out. There's always more research. Always. EIL, the most challenging part of this episode. If you could sum up this conversation in one or two sentences, what would it be? I think I would sum it up, first of all, by setting the context that informal settlements are not temporary in countries that are experiencing rapid urbanization. And actually predictions are that they're going to grow immensely in the next 10 to 30 years. And so if we want to think about planning questions related to service provision, like what are pedestrian activity patterns, which is something, by the way, that planner study in great detail in formal areas, it's essential that we also study them in informal settlements in order to ensure well-being and quality of life in these neighborhoods. And by looking at these patterns, we have the chance to better inform all kinds of different decisions that really impact people's daily life, the way their habits and their enjoyment of their everyday experience. Great episode, EIL. Thank you very much. Yeah, thank you so much for having me. It was really great to have this opportunity. Sure. And for those who are watching us on YouTube, on the description on this video, you can find all the resources, all the information about the study that EIL and I have just discussed. You can also find the links to our podcast platforms for our Twitter account, for our newsletter to stay in touch with our episodes. And we'll see you in our next episode.