016 - The future of evacuation modelling with Enrico Ronchi and Ruggiero Lovreglio

Wojciech Wegrzynski: 0:00

Hi there, everybody. Hello and welcome to the Fire Science Show session 16, a nice round number two, the power of four. Nice to have you here today. We're going to discuss of evacuation modeling. A part of fire science engineers life, that's always there. If you're doing any type of analysis, you're usually relating them to the human safety. And the human safety factor is usually evaluated by the means of modeling the evacuations, egress phenomenon. And today I have two guests. Both of them are renowned experts in modeling evacuation, in researching the various aspects of how people escape buildings. The decision processes. The choices they may come, and the factors that influence the whole process. Both are Italians for some reason. I don't know if it's, uh, it's a thing, but it seems Italians are powerful in this field of fire science. Um, they're both connected. One was a supervisor. One was a student. Now they're both professors. They're both young stars of fire engineering. So, yeah, please join me in welcoming professor and Enrico Ronchi from Lund University and professor Ruggiero Lovreglio also known as Rino from Massey University in New Zealand. The idea for this episode came to me when I was preparing a presentation for Polish experts on. It's current state and the future of evacuation modeling. Back then, I sent an email to the Rino and, and Enrico to ask them, what do they feel about the future of evacuation modeling? And I've received some really cool and vibrant answers, and then immediately ask them to maybe share them with you in the here. And here we are discussing all these things from. Simulations in buildings to virtual reality and augmented reality, data gathering and other concepts of evacuation modeling up to wildfire evacuation. So it's jam packed with action. I hope you really enjoy this one. So yeah, without further ado lets spin the intro and jump into the episode. Hello, everybody. Welcome to Fire Science Show. I'm today with not one but two fantastic experts on evacuation modeling with me today. Professor Enrico Ronchi from Lund university in Sweden. Hi Enrico.

Enrico Ronchi: 2:54

Hi, Hi,Wojciech. Nice to see you and hear you.

Wojciech Wegrzynski: 2:59

Great to see you as well. And from the almost exactly opposite side of the globe, Professor Ruggiero Lovreglio from Massey university in New Zealand. Good evening Rino.

Rino Lovreglio: 3:09

Hello everyone. Nice to be here and thank you for inviting me.

Wojciech Wegrzynski: 3:13

Thanks for taking the invite, you guys. And I really appreciate that. I've invited you because I wanted to discuss with someone, the future of evacuation modeling and how it's being used and how it can be used in the future for better. And I was thinking about an expert and then figured it out. I know two experts and they're both Italian. They're both nice. So should work out here. You are. If, if it doesn't work, I'll send you both a pineapple pizza

Enrico Ronchi: 3:39

Ouch. yeah

Rino Lovreglio: 3:42

Hurts.

Wojciech Wegrzynski: 3:43

Okay. So before we, we, catch up for the podcast. I asked you about some ideas about the future of evacuation modeling, but, let's start with how modeling is done today. And, I just want to ask you this cheesy question. what do you think is the role of evacuation modeling in current fire safety engineering for you Enrico? Maybe you can start.

Enrico Ronchi: 4:05

I see evacuation modeling as a useful tool, within performance-based design, in terms of, yeah, we use it to calculate RSET, but not only nowadays, because it also gave us the opportunity to identify a critical bottlenecks in, buildings. So even to spot, if there are some potential issues in a given design and also .The part that is very powerful is the one that relates to the fact that we can simulate several "what-if?" scenarios. These put us in a great position, especially when we want to consider it scenarios indeed like fires. , In which we don't really often know exactly what will happen, but we can guess a set of reasonable set of scenarios about what could happen. I see it as a very powerful, a useful tool, but there is a, but there's always a room for improvement, uh, otherwise, we will all, uh, let's say, be happier and never improve if we don't think about, these in this way. But, at the moment, I will say it's still a useful tool, for performance-based design and not only, so also to have a rough idea on how building works in terms of design when you put people in it and you put different behavioral scenarios in

Rino Lovreglio: 5:17

Yeah. I want to add that box is also good for people that work in the industry because they are not automated models that you just tell him, this is the building. and Do the work for me. So we still need brains and engineers capable to understand what are the parameters to put inside those models to try to mimic as much as we can the reality. And that's why most of the course that you will find around the world. They don't teach you just to click the bottoms. And that's the things that probably we say all, to all our students that probably your cousin that is a geek in computers will be much more capable to use any software that is around the world, but you are going to get paid in the future because you are capable to understand what the theory of human behavior in fire. And try to do your best in uh, integrated those theory, and the right numbers as input of those model. Because the main issues is that at this stage those model are heavily rely on the inputs that the user put.

Wojciech Wegrzynski: 6:15

Yeah. So you're saying by understanding how to build the model, you understand the human aspect better. And then in general, it gives you closer to a safer building, like electric vehicle forces you to save energy by driving more calmly in a way. that's a good one. For me, as someone I'm not I'm researcher in this, field or my research overlaps with, with the evacuation sometimes, but it's not my, my principle and area of research, but I'm using lots of evacuation modeling in my engineering practice and for us, sadly, it's an RSET calculator usually. Well, in ITB, we would do this bottleneck scan that Enrico mentioned. I think this is very powerful because in generally buildings work, if you build a building to a code and you have this evacuation paths that are required, at least by the Polish codes, it usually leads to more or less a safe building. Not, perfect, but not horrible either. But this, this bottlenecks can, can be very limiting for the evacuation process. When you have the complex geometry of the building it's a perfect tool to scan for that. However, it takes some skill and knowledge and especially the constants of the behavior, or speed, which we'll touch in the second are very important to be defined. Rino, I know you also, use a lot of, this modeling or this approaches connected with, virtual and augmented realities to study the behaviors. And I think it's something that opened up, lately, a lot. So how does this, interplay with, with the evacuation modeling?

Rino Lovreglio: 7:43

We must say that most of the work done in virtual reality, wasn't initiated by me. There are people that have done much more on this just to mention Max Kinateder did a full PhD student in integrating virtual reality. And also Enrico and Daniel Nillson did the quite a, lot of application of virtual reality. So I basically, they prepared the road for me. I basically followed the path that was, already, defined by them. And as you can imagine, virtual reality give you the possibility to build evacuation systems that don't exist, or to test a new concept of evacuation system and try to see if they works. And uh, you can see for instance, the, one of the most relevant paper on this is the paper written by Enrico on the tunnel exits. If you think in the past, before having this kind of study the poor Daniel Nillson who was supposed to manufactured the system, and bring people stop a tunnel and bring people in a tunnel You can imagine from a logistic point of view was a nightmare. But at the time it was also a nightmare to have virtual reality back then, because he was incredibly expensive. It was like something unbelievably expensive. And now it's becoming so cheap. That it's becoming quite easy in the labs to have it and use it for testing evacuation system and also to push a bit the boundaries to study a bit social interaction. And for instance, Max Kinateder did a lot of good studies on social interaction. In my case, I'm a, I belong more from the path of, a I'm not up psychologist. I'm more oriented by the transportation guys in which we try to put, many variable and try to see how many factors can affect the, the behavior of people. And virtual reality gave me the possibility to experiment what will be the, the, the the decision-making of different evacuee considering different combination of variables, and there was something, that it wasn't possible even. for my PhD. In fact, in my PhD, I had to use non-immersive technology, but then now since price is dropping and, buying VR headset it became quite easy for everyone allowed me to do much more immersive study, to, to go deeper and try to understand all different combination of variables and factors can affect the decision-making of people. And these is something awesome.

Wojciech Wegrzynski: 10:03

Bottom level, you can buy a cardboard and put your smartphone in it, and you have some sort of virtual reality, which is actually not that horrible. You were referring to the, um, dissuasive exits paper if I, if I'm correct, right?

Rino Lovreglio: 10:15

Uh, that's one of them. The one that I was referring there was about, it's a paper in Fire Technology. on no,

Enrico Ronchi: 10:20

Flashing lights is the one on

Wojciech Wegrzynski: 10:23

Yeah. Flashing lights. Okay. Yeah.

Rino Lovreglio: 10:24

And the good things that the followup of the paper. led by Enrico. It was like to prove that, the experiment done in a cave that, he took. I think the cave is still incredibly expensive and the same experiment In in a really cheap headset, ended up pretty much with the same results. So at least for testing uh, fire safety system, you can even use pretty cheap devices that might cost you 1000, 2000 Us dollar, or even less, and still you can do really cool things. And Enrico probably can tell you more about that because he was leading.

Enrico Ronchi: 10:58

Yeah. I mean, the idea there was that for the fire safety engineering purposes, when it comes to studying evacuation system, we, we originally did some studies in a cave that we have at Lund university. The cave is basically a full scale room in which we have projectors and we deploy VR there having some sort of ultrasound system that tracks your movement in VR. Those are very expensive labs. And this is what was traditionally used. Let's say for more than 20 years to do VR experiments at the same type of lab that they have for instance, at university of Wurzburg the one that Rino mentioned with Max Kinateder. But then what we did after doing this experiment to test how a certain type of evauation system worked. We thought, okay, what will we get if we use a much cheaper, VR solution? So if we use some sort of a headset, and we went for the cheapest that we could find on the market pretty much, because we went for a cardboard with a phone inside, and I mean, If we look a certain type of variables that, do not get so much impact from the interaction with the environment, but it's more like a visual system that you are interacting with. We saw the results were pretty consistent, so it was very encouraging to see that basically you could get reasonable understanding of what will happen in a fire evacuation scenario, uh, using a very cheap virtual reality system. So that was, uh, the idea and the, and again, we have to be careful in generalizing these things because it might be that there are, uh, scenarios which require more interaction for which, uh, there is an advantage in using a more immersive system. In most of the engineering problems that we have for fire evacuation design, we often have to decide between a different evacuation systems or different design choices. So things like that affect actual layout of the building or that affect the actual installations that we put in the building. So in those cases, often it is possible to have a quite decent grasp on how people will interact with those by, uh, using, fairly cheap VR. And, and again, this is something that, uh, from being expensive and rich labs, a university now they are a consumer products that basically pretty much anyone can afford.

Wojciech Wegrzynski: 13:03

Do you have to go through ethics committee for these experiments? Or how does that path look?

Enrico Ronchi: 13:08

We have to go through ethics committees, mostly every country and every place in the world have different rules, but in Sweden, we have a national committee that evaluates let's say the risks and the possible issues, for instance, with the pressure on the other protection and so on of the experiments that we do. So when we deal with human subjects, basically we had to go through, uh, ethics. But at the end of the day, the type of experiments that we do, they're not too different from what is a gaming experience. Actually, I will say that it's much more mild, it's much milder compared to certain games that we have nowadays.

Wojciech Wegrzynski: 13:43

Yeah.

Enrico Ronchi: 13:44

To run over people or shoot at people like in certain games. So it's more like, experiencing a fire scenario and evacuation scenario, but again, we have procedures in place for making sure to minimize risks because you know. There are still a portion of people that they may get dizzy when they use VR, or they may feel nausea and so on.

Wojciech Wegrzynski: 14:05

I've asked that because I know it's a hell to get a fire experiment approved, like exposing people to real smoke in the real, smoking conditions. I once talked with Daniel Nillson and then he told me he has horrible issues getting an experiment like subway station approved because of, of his willingness to use the real people. And then these experiments with real people are also, usually for some reason, limited to, you know, a population of healthy students. And while your VR set, when it's cheap and easy to deploy, it's also something that could be used to cross cultural boundaries, you could run the same experiment in multiple places of the world to expose different populations with different backgrounds. And, you know, in a way our evacuation modeling, the fundamental diagrams, the pre evacuation times are, let's say like very white European population bias, right. Because it's a, it's been Lund it's been Greenwich it's been generally European, units that have been developing that. So, um, yeah, I wonder what do you think is, the future of this, prescriptive approaches? Like the, the fundamental diagram, the, speed versus density plots, the pre evacuation distributions are the things that currently probably determine, the output of evacuation modeling more than the physics of the model itself, or often in a comparable way. Yeah.

Rino Lovreglio: 15:25

I got, that we are trying to open here, the can of worms, because, because there are a lot of, uh, standard curves that have been using for years that have been generating for specific types of population. And there is a lot of papers that we are bringing in the field and showing that, those curves can be probably too conservative for some cases, but in other cases won't be conservative at all. So that means that we might get in trouble. If you use. We expect that those curves represent the real behavior the real trends of pedestrian dynamic, and then they are not exactly that. And, the issues that we have now is that, we try to find. the, If you don't use the standard curves, like . for instance, the SFPE curves, we try to find curves that we're collecting in a population that were as similar as possible as the population that we are. we might expect in in the building that we are designing. And this is still a issue because that means that we need to have incredible different variation of those curves And interesting work that was initiated by Daniel Nillson Uh, Karen Boyce Steve Gwynne and Denise can't remember her, surname sorry, Denise ( McGrath) It was like, trying to to find the really fundamental, fundamental diagrams So fundamental squared diagrams In which you need, just to define that physio mantric characteristic of people. And then the fundamental diagrams will show up as our input. And that's one of the things that I was telling you. like You can rely on, defined inputs, of models. like, just because this is the standard curve. I need to use that one at this stage, you need to even find the fundamental curves that represented are closer, to, to your case study. And for instance, if someone is trying to do the job now, in my goal on there, chapter 64, I guess, of the SFPE chapter it is on engineering data written the, by Steve Gwynne and Karen Boyce. And in that case, there is a huge collection of different kinds of, speed and fundamental relationship that you can find extracted from different experimental condition or real case,

Wojciech Wegrzynski: 17:38

And then there's also your paper from NIST that, that covers a lot of specific

Rino Lovreglio: 17:42

uh, pre-evacuation. I try to do something to help people to identify the right

Wojciech Wegrzynski: 17:46

You helped me thank you.

Rino Lovreglio: 17:49

But the, probably some sometime when you publish this kind of paper, you feel like I'm helping, or I'm creating a bomb? Because if those curves and distribution are misused, you're basically creating a damage for society instead of helping society. And that's the big risk. And that's why we put a lot of statement in the paper that you're referring about, that pre-evacuation distribution. About be careful about how you select, because it might not really represent what you're trying to represent with your design. And, uh, to move away from this kind of approach with Enrico, we try to write and study a lot of other alternative approach in which we actually try to model the decision-making in terms of pre-evacuation instead of relying the on curves, because those can be the results of our agents based modeling approach rather than, than, uh, defining input. The problem is then now we use pre-evacuation as I input, but the dream will be that this will be the outcome of a model.

Wojciech Wegrzynski: 18:50

So you just define the agent that it exists and it's in this place. And whatever happens since you press play is it's the model, not

Enrico Ronchi: 18:58

yup yup

Rino Lovreglio: 18:59

On the other wave. The pre-evacuation is that these might make a fire engineer useless at some point, but that's kind of in an ideal world. If we manage to develop models that are capable to predict things, without smart users, at some point we will get rid of foreign engineers, but I think this is probably, we will. see these in two, three generations.

Wojciech Wegrzynski: 19:20

Well I'm not sure if that be a bad thing. I think you would not get rid of fire engineers because we are generating fire problems quicker than we can solve them. So there's always a job for fire engineers. Our, job is very secure in that manner. And on the other hand, there's always not enough fire engineers

Rino Lovreglio: 19:37

I hope I won't receive any call because sometime I teach my student construction, students say that AI is gonna take over replace and sometime after the classes, there is some students that approached me and said, Rino, I'm really wasting my money. Because my parents are spending so much money in my degree. And then I make sure that they understand that it wasn't basically a joke.

Wojciech Wegrzynski: 19:57

What's your, what's your, taken on fundamental diagram?

Enrico Ronchi: 20:00

I mean, it's, you know, it's controversial in itself, even if there are fundamental for fundamental diagrams, because

Wojciech Wegrzynski: 20:06

sorry. Are they fundamental because of physics or are there fundamental because we have to use them, right?

Enrico Ronchi: 20:11

I, I that's, that's kind of the feeling sometimes that we don't have anything better at the moment. So we have to go on with them. but there is also some aspects that I want to bring up that is a bit linked to what, Rino was mentioning before, that, we are designing buildings today that will be used for several years from now. Yeah. So this means that we cannot really rely necessarily on, design that is based on the population of today or, or even worse population, uh, data that has been collected many years ago because the population demographics will change. We know that we're getting older and the prevalence of disabilities in the population is increasing. So we need to find ways to change this paradigm. find a way to design buildings, projecting what will be the population, in those buildings, not just looking at what it is today, or even worse, relying exclusively on what data we have from the past. Data that they've been collecting the seventies, eighties and so on. So we, really need to change that kind of a mindset that, okay, it's good enough, to design something with what we have so far, we have to try to push the boundaries, especially as researchers, we have to try to see how we can solve this problem. One thing in this aspect that I want to bring up is. regarding the, let's say variability of possible population that you can can have in a building to me, uh, recently I've been involved in quite some work looking for instance, at different types of disabilities and or what type of functional limitations people may have in a building. And I will say in, in fire safety, we are very much behind compared to other fields like accessibility or even looking at the medical science. So, you know, often you read the fire code. And you read something like, okay, you should take a profit provisions for people with disabilities, but you know, there is so many possible disabilities, so many possible situations and scenarios that we should have to take into consideration. So one important step that research should do is to try to categorize in a much more systematic way the populations that we have, uh, based on the functional limitations that people can have. so. What are their characteristics, what are possible limitation they can affect egress and so on. So that's an important step, especially if we look at the future, because we know that we are getting older and older on average is population, which means that the also where we're getting more and more people that have, more issues, during evacuation. So I think this is an important topic that paradoxically is not being so much investigated to me. I mean, I see as one of those things that we should really put a lot of emphasis in research, because it's going to be more and more important,

Wojciech Wegrzynski: 22:56

This is also something I've been, I am asked about in, let's say one out of three buildings that they do, that, how did you account for disabled people in your evacuation modeling? And I usually say I've hidden them in my safety margin, you know, because I don't really feel, I have a powerful way to account for them today. I could technically, yeah. I assume a small portion of the population in my model is, disabled and has completely different. Let's say movement or pre decision time characteristic than the rest of my population. And then place this few people randomly around the building. But then we come to the problem of how this exotic point affects the outcome of the whole simulation. And when my population is uniform, convergence fairly quickly on the random distribution of people in my building, because if I ran five, maybe 10 simulations, it usually will converge to a value. But when you have these exotic points, basically few people among a hundred or a thousand of in your building, you would probably have to run a thousand simulations to do find this effect. Then eventually figure out the bottlenecks scenario. So it seems like very, costly procedure. so I prefer to hide behind the safety margins. I'm not sure if I'm doing it right.

Enrico Ronchi: 24:11

I guess that's a one way of doing it, but we are still, in a situation in which even if we want to model explicitly, the wide range of possible disabilities and how can they affect the evacuation? It's very hard, because we know relatively little, so often you see, for instance, in buildings, especially more complex buildings that you have dedicated solutions, for people with disabilities. But as a general thought also for that, I think we have to move away from that concept. I know that this is practically often implemented, but the ideal scenario would be to find a solution. Equally works for everyone. And if you, if you're using a solution, the work for people with disabilities generally works also very well for, call the standard? The, uh, or whatever we want to call it, the adult population, uh, whatever standard means, because again, the standard actually, we're going to get to the point that we have to change. What is the concept of standard population? Because this is what it is about here. The future building we'll have more and more, let's say variability. that's why I think it's a reasonable approach, what you're doing Wojciech nowadays. But if we look as research, we have to look at the future and that's something that we should change. At least we should try to cut up the rice better, the populations as it is now. and then the projection of the future is the.

Wojciech Wegrzynski: 25:30

Think it's also now with the tools becoming available like this, immersive setups or just virtual reality without exposing these people to do true harm, allows you to really quantify things that you could never do in the past. Like you would not put, I mean, it's hard to put students into smoky tunnel and think about trying to do experiments with disabled people on the same setup. It's I think it would not be possible. And yet today, you, you may have, uh, chances to quantify this. I also, I also think like, to what extent we can capture this data from the drills or real fire accidents, like the buildings today are armed, with CCTV cameras. So technically if you have a shopping mall evacuated, if you have a cinema evacuated, you can pretty much figure out with machine learning, what happened, where people did move. Right. So how do you feel this sources of, of data can, can play around in, in making what you just said before.

Enrico Ronchi: 26:29

Yeah. Well, they will definitely help. I will say, but yeah, there is all sorts of problems that are more related to privacy issues because, you know, then we enter a territory to like, to which extent are we even allowed to use CCTV cameras or whatever other data I know for instance, Rino worked a lot with GPS data, you know, or, uh, even, you know, there are people that use the traces of your phone in a building The big issue here is not the technical feasibility, because there is a lot of work already been done in this, it's are we even allowed to do this? So for the good of science, I will argue that we should, but, uh, sometimes Oh, go to an ethical committee and explain that you want to film everyone in a building for one year, until there something happens that it's interesting for you as researcher. And you can check whoever is in that building all the time. I mean, it's hard to convince both the ethical committee and even the owners of the buildings very often, uh, that we do it for the good of science.

Rino Lovreglio: 27:30

If I can expand on that sometime paradoxically I've been recording drills in some buildings in which there were CCTV videos, but they told me its easier if you put, I mean, from a regulation point of view, or this institution, it was much easier to get the approval, to put my own mini cameras and record the drill, rather than, then using the follow the process to get the video through all the process that they had in place to to access the video through the CCTV cameras. Sometime it's incredibly complex, the process that you need to. follow to get access to video that are normally recorded. And I had the same issues here in my university to get access to the CCTV recording of drills, to go through our really, really heavy protocol and explanation of why and how the data will have been used And, uh, sometime it became really, really challenging to even to get access of recording of even tragedies of a real accident.

Wojciech Wegrzynski: 28:31

so probably easier to get them from YouTube then from...

Enrico Ronchi: 28:33

yeah

Rino Lovreglio: 28:34

sometime it happens that you find papers that are based on social media. Videos.

Enrico Ronchi: 28:38

There is a bit of mismatch there. I have to say on a, how much care and attention there is in research about data protection and what you see out there. I mean, it's like if you see a TV show and you see like they do a social experiments, many of those things will never get approved by an ethical committee, but it's a TV show. So the get away with it. So it's kind of, I mean, as a research, sometimes it's strange. It feels really strange. There is a different social acceptance between what are you allowed to do as a researcher? And in theory, as a researcher, you are the one that know how to do these things according to ethical standards and what the rest of the world can do, because pretty much as it's for entertainment or whatever, or the reason, you know, because I mean, we could, we could open now a very long discussion about privacy and how much we give away our data to the world, without even knowing sometimes versus how much we are sometimes restrictive to give our data to researchers that will actually make a good use out of it. They will not try to sell us something.

Wojciech Wegrzynski: 29:42

That's in a way horrible, because what you want the data for is to essentially make the building safer for exactly these people. And when you go into their privacy, you go for this particular part of privacy, which is a behavior choices, movement, patterns, not why Karen went to shop to buy this particular product. No, you want to know what, what was in her head to choose the roads that she picked up. So you can design it in a way that in a fire she'll have a better chances to escape. Me as a fire engineer, we usually like, people don't know how much time in engineering it's takes to buy them one minute of a time. Like how much work I have to put to extend my available time in, in the car park for one minute. And then when you see a fire and, uh, this one minute was well spent for creating a new video for YouTube. And it's kind of painful, you know, but, , Here I, I jumped in into another aspect, that comes from eighties, 1970s, like fire safety Is a field built on tragedies. And in evacuation, we had this huge fire in Las Vegas then was the World Trade Center. These were events that triggered like breakthrough or leap through, uh, research projects. And then there was no cell phones. People were working in a completely different way. If you call your friend to meet him on Friday in the Plaza at noon would be there without sending seven text messages. And and if he was not, it was not a huge issue either. So, uh, you would just live on with your life. And, and today the, the society and individual behavior is like completely different. Not to mention the shape of a human. I'm definitely not fitting the anthropogenic average of a, of a male in my age, uh, from the, from the seventies, I have a big, much bigger moment of inertia. And, but this, this, you know, we, we base our modeling on this assumptions. So can we leap forward? And you also said something very important that is it enough to leap forward to today or we have to leap forward to tomorrow because ultimately if you design the building today, it will be used in 50 years. And who will be using the building then some Jetson people. how do you think we can, like make this leap in a reasonable way? Or can we even make it.

Rino Lovreglio: 32:08

Somehow I think that in the future, we will have easier way to communicate with people inside the building, because our building out slowly, I should say really slowly getting smarter and smarter. And one of the main issues that we have, uh, and we used to have in the past in many tragedy is not because we had a lot of casualties. So people getting harmed, not because they were panicking, but because they didn't have the right information at the right time. And my hope for the future is that those building and all the technology that we have available will allow us to, I have a right information much quicker and help us to make the right choice. And this gonna, I feel like the the, the building are somehow is gonna, are going to become safer because this channel of information is going to get better and better. And if you know already, where is the fire? and You have the possibility to communicate to the building occupants where they need to go and what they need to do in real time. They will make a big, big change in, in the future. And, you were mentioning before augmented reality is something that now we are just conceptualizing about. We are doing. Small experiment, but you you can really imagine that from here, to 10, 20 years, holograms will be in a part of our life. That will be like a smartphone. Now, now we are looking at a smart man thinking how the hell we've managed to to live without a smartphone, I, all the old people, were capable to, to find a place without Google Maps or meet each other. At the right time. Or even when they were traveling together with two different cars or they managed to not to get lost

Wojciech Wegrzynski: 33:52

Yeah.

Rino Lovreglio: 33:53

For my generation is like, that was crazy instead not, you find a way, And, and I think that augmented reality and holograms are going to make our life safer and easier if we use it in the right way, because we can use all of them to, to guide people, to communicate with people, in a more dynamic way. If you think now we have a dynamics that we don't have much dynamic system that lead you towards the best exit. We have just a static life that our needs need to be there just to tick a box most of the time, because the code tell you to do. it say And where we put most of the signs seeing now of course, on the top, because you need to see them, but the small goes on the top. So there is a lot of things that you do it, for checking the box, but from a human behavior perspective and performance perspective, you know, already that it's not going to work

Wojciech Wegrzynski: 34:49

But if you go into dynamic systems, um, there are dynamic in, let's say transferring information, but the scenarios for this transfer are still static. So if you have a fire at the seventh floor, they will trigger to this area. If, if not, they will trigger to other area. And from my experience, the ones that I've met, they were like locally connected to sensors. So if there's a fire or behind the door, it will show X, but it does not like really think about what's the optimal way. And we also are arming our buildings with, with sensors, with a lot of devices. And it's a theme on the podcast, whoever I speak with, they see this immense value of, of arming the building up. So can we use this let's say smartness of a building to maybe guide the evacuation process and if we can, should we, or can we, um, include that in our modeling? Because if ultimately like I can imagine that arming a building with a smart system would make it safe. But should I, in my modeling, take it into account that the building is smart is safer and evacuation time is quicker, or should I go conservative?

Rino Lovreglio: 36:04

At this stage we'll go conservative, of course. But if you plan for a conservative, scenario, and then you manage to bring in a building technology that is going to make it much more efficient from a safety point of view, that means that it's going to be even better than what you design

Wojciech Wegrzynski: 36:20

if I go conservative in my modeling, how do I explain to my client why they should put additional money on the sensors? If they don't benefit? Eniko you've mentioned smart buildings in the

Rino Lovreglio: 36:32

I see what you mean

Enrico Ronchi: 36:33

Yeah I understand the problem because then we go always to the bottom line is engineers, is this

Wojciech Wegrzynski: 36:38

And, and sorry. if you go conservative, you can just say, okay, there's 5,000 people in my building, the speed they can exit doors is one person per meter per second. So it's going to take this amount of seconds. End of the calculation. Why, why model anything else?

Enrico Ronchi: 36:51

There are two aspects here. One aspect is something that, unfortunately in most cases, safety is still seen, a box that we need to tick.

Wojciech Wegrzynski: 37:00

Compliance not safety

Enrico Ronchi: 37:01

Exactly. So, in an ideal world, safety will be, for instance, if you go to hotel, safety will be as much rated as comfort or other things that will make you choose one over another. But unfortunately the reality is that very often, this is not the case. So you need to meet the minimum bar. Then if you are exceeding that minimum, let's say very little people care. And often the clients don't want to even pay for exceeding that not in all the cases, but in most of the cases. What I will see as an ideal let's say an ideal set up to really exploit what we have nowadays will be something like that. Connects the sensors with a modeling system. Smarter in a way. So, and there are a few attempts of research groups that have tried to do that. I know that the, the group in Greenwich is just started

Wojciech Wegrzynski: 37:50

There's a group in Poland

Enrico Ronchi: 37:51

Well, basically you, you have, building the district with sensors. This information is sent to a simulator. The simulator also knows where the people currently are knows where the threat is, and then run a set of what if scenarios and analyze what is the best route to indicate to people to get the quickest and the safest, uh, evacuation. I mean, it sounds like scifi nowadays, but we're actually not that far from getting there. I mean, there are a few things to solve concerning. Let's say a reliability of such systems and, you know, things have to legislation, things have to hold during a fire, everything, the communication

Wojciech Wegrzynski: 38:28

two directional communication because today it's simple. If your fire sensor detects the fire and dies it's okay. Because it did its job. Yeah. But here, if your, um, human counting mechanism fails at some point

Enrico Ronchi: 38:42

Fails Yeah And then you are in trouble because you don't know anymore which direction to indicate people. You don't want to point them towards the threat and so on. It is not as much far in the future, the way I see it, but there is quite some work to be done yet. Not simply on the technical side, but on the redundancy set up that you have and the reliability of such type of system, because that will be really ideal. So like to have really smart building that makes smart people, if you know what. That's the way forward, the way I see it, because all the systems that we have now tested like a, you know, you see dissuasive signage and the active signage. And so on this, you know, there is research that tells you even how to design, to provide the correct information, uh, to people. So, yeah. We have some ground research already in there, but what is missing is like, okay, we need really some very large company with a lot of money to put in, into actually making something reliable and resilient

Rino Lovreglio: 39:43

I can see the, For instance, here in the west island of New Zealand, it's called Australia Some Australia might get her with this joke... We have even a company that is producing those of dissuasive sign. But the problem is that you can develop all the manufacturing, all the system that you want, but it's from a legislation point of view, we need to address what's is going to be accountable for responsibilities if those system fails. So there is, a, it's not just a matter of developing the technology, but also rethink about the way. We share responsibility in terms of who is accountable for what, to do after a disaster.

Wojciech Wegrzynski: 40:24

Developing the technology and creating the algorithms is the easy part. Making a world use them is, is the hard part. We've learned it the hard way in smoke control. If you come up today with the new smoke control solution, that is much better than the previous ones does not matter. As long as the previous ones are compliant and there is no economic incentive in using the new system, it will not be adopted. While any the smallest economic incentive, like this solution makes it's cheaper than immediately a landslide starts where people put money, investing in these technologies, improving them. Eventually you get with very optimized solution beyond the point where, where you cannot optimize anymore. And this solution gets there. It gets implemented in the buildings. And this is why I've asked you to know that, uh, how can we account for having these advanced solutions? Because if I don't show incentive to my stakeholders, they will. Yeah. That's, that's pretty cool. But I have a yacht to buy

Enrico Ronchi: 41:28

If I can add something, Wojciech. One other thing that you can do is that we should not have this kind of smart sensors just for safety. So if we use systems for multipurpose, so I make an example, like a couple of years ago, I was involved in research problems with the CERN in Switzerland and the, these guys, they have all sorts of technology. And as you can imagine, and for instance, they use robots for maintenance of their tunnels. So they have a railing system with robots basically, or raiding there. so I said, okay, you have this there. So why we don't use it also for safety purposes. So why we don't rig this with some sort of signage, they move around, chase the people and indicate the way. And I mean, it's, again, it sounds like scifi, but. this idea. and we have even tested in VR to try to see how it works and it actually does work. So because the buildings are getting smarter and smarter, so they have many things in there already. So if you're using something that is already in there, or you're proposing a system that can be exploited, not just for safety, but also for other purposes. I think that's the winning argument for a client saying, okay, look, you are putting the scene, but you're not going to use it only for safety. You're going to use it maybe also to facilitate whatever other let's say, purpose that you have in the building. That that should be a good,

Wojciech Wegrzynski: 42:48

yeah Xinyan in his episode, he was facing the same problems as we now in this talk, because he mentioned about HVC sensors that are used in the building, but not used for safety. For example, they are used to control the climate in the building. So we have to measure temperature, humidity, and some other parameters of there, which could be useful for, for fire when it's developing in the building. And, and I think he, he, he mentioned something very smart High-value high-risk buildings such as tunnels. Big museums could be great playground for, this approaches, in modern tunnels, you already have very powerful CCTV systems With accident detection. So they know if a car stopped, they know if a person exited the vehicle, you have this two way communication, at least in the tunnels in Poland, you would have some sort of voice alarm system and a microphone to communicate with the people inside the tunnel. So, so these buildings could be like a testing ground to figure out this, uh, solutions. And if they work then maybe we can show how they help in buildings. And if we do that, maybe we can scale up to let's say buildings in general. That's good route ahead. And from, from this discussion, I'm kind of surprised it did not come up yet and that we, do a little bit better job connecting, like, CFD and evacuation modeling, you know, the viability of ASET-RSET concept, or, or this is another kind of surstromming you don't want to open. Yeah. I find that this a fascinating, because it's like the dream of evacuation modelers or since evacuation modeling was born to connect both. And it gets rid of ASET-RSET

Enrico Ronchi: 44:31

And two way so that the actions of people affect, the fire. That's the, the, the real missing part at the

Rino Lovreglio: 44:39

Okay

Wojciech Wegrzynski: 44:40

that That's a good one

Rino Lovreglio: 44:41

this Can you

Enrico Ronchi: 44:41

I mean I I, can start it at the moment, most of the model they do some sort of coupling. They take the fire, uh, outputs and they put it in evacuation model and they try to affect the behavior with all sorts of limitation that we are aware of. For instance, well, I mean the one that is the biggest and more evident to me is the CFD calculates visibility and how the system Wojciech, I don't have to explain this to you, I guess. Uh, but you know, for me that I am the other end with the, RSET. I mean, this is, sounds like it should be priority one for any CFD model to improve the visibility sub models in CFD, because that's what really matters for evacuation. Everything else is kind of. Yeah. I keep seeing afterwards that the visibility criteria is the one that screw up everything generally

Wojciech Wegrzynski: 45:31

You cannot be tripled dead I mean, you die because you lost path is you're down Not that doesn't matter if the temperature killed you twice

Enrico Ronchi: 45:41

Exactly. Exactly So, I mean, th the ideal thing would be actually to have a real integrated tool in which the actions of people down the line have a direct impact also on the fire calculations, because you know, better than me, that if you open the door, close the door or do anything in the building that can have an impact on the configuration of the building layout. That's like you change boundary conditions. it's another scenario

Wojciech Wegrzynski: 46:08

This is, Fantastic. That you've mentioned because it it blows my mind how weare not doing that. And even in practice, we doing the complete opposite. I've done CFD simulations because I was forced to, unwillingly, to show that when you close the door to compartment, your corridor is free of smoke. I'm like, I seriously, I don't need to do CFD simulation to tell you that if you close the door, it's going to be okay. Oh yeah. But we need this scenario, the fire expert wants but why, why would I simulate a fire in the compartment where I closed the door before the fire starts, but they wanted it. And in here, what you mentioned, if you follow a person and you know, they left, there's a, let's say 50% chance they left the door open or closed in, in the room. That's two completely different CFD simulations of that, of that scenario. So, um, you you're right. Two way integration is something

Enrico Ronchi: 47:03

And also with some sort of a, let's say probabilistic approach, because you will not know, it's not only about if they leave a door open when it's already said probably it's also when and what episode it's like in this case, you could go directly for probabilistic approach in which you could set up a series of possible scenarios and see the, the consequences on the global, uh, probability of having something to happen and what is the impact

Wojciech Wegrzynski: 47:32

Yeah, Ruggiero you did something like that with this evacuation of a musical festival chemical accident near it

Rino Lovreglio: 47:37

Yeah this wonderful idea. No it was Enrico's idea yeah. The real story sad than real stories that I was writing my thesis, my PhD. thesis, and then I was getting mad. So, Enrico told me, I give you something to work. But still the prominent was still the same that affect the fire simulation. And I'm not entirely sure that this is right, because if you think like a crowd moving, across this, moving here is creating a perturbation in the, in the field, uh, that where you're simulating, because when you the simulation, uh, for instance, that we use in that case were done in CFD with the best that we could do consider the computational power that we had, and it was done by the team in Ghent and of course it wasn't considered in the perturbation of moment of people in in the space

Wojciech Wegrzynski: 48:32

Calculating doses and stuff

Rino Lovreglio: 48:34

Yeah The things that we managed to do is to get the simulation done in CFD, Enrico. Did the simulation about the evacuation with, one of his students. And my work was just right to connect, and calculate what was the dose of inhalated, uh, gases by, by those people. But the, the coupling once again was one direction. It wasn't bidirectional.

Enrico Ronchi: 48:59

Mean, that was also an outdoor scenario, so it's a bit less critical, I will say. but I mean, in principle I still see a clear gap there. I've seen only very limited attempts to do them and mostly in a very premature research stage. So, you know, okay. I've seen a PhD. I think there was somebody doing a PhD on this topic in Greenwich at some point, but you know, then I didn't see anything else. So it's really hard also sometimes to convince a, let's say research funder, then what we're doing at the moment is still very limited because you say, okay, but we have buildings until now they stand. There is a clear the app they're on because this is something also in which we have a limited knowledge from the human behavior perspective. You know, we general understanding of what action people can do, like things as what you mentioned, like leaving a door open or not even a door open and what is the probability and so on. Uh, but you know, before putting this into a model, we will have to do a much more systematic investigation of what could happen.

Wojciech Wegrzynski: 50:00

Okay. Let's move. Let's move back a bit. One step, let's take this two dimensional coupling let's, let's go back to one dimensional coupling. So only the effects of the fire are simulated on their evacuees, but let's go back to outdoor and then let's talk wildfire evacuation, for example, where you evacuate, not not a building, but a, an area filled with people. And here you come to similar choices, actually that you would have evacuating building because it's, again, a topic of the speed at which you can travel. The route choice model and pre evacuation time is just the whole the three elements are completely different so

Enrico Ronchi: 50:35

Yeah I mean it's a, the may not that late, we are there is that we don't evacuate only on foot. That's the thing. So we have the traffic, modeling part, which is a completely different world I have to say it's much bigger than the pedestrian modeling or it's much more investigated. That is much more, uh, critical, massive research. I should have looked into this mostly for general use Not necessarily evacuation, but that has been also a lot of research. I know Rino, you have been reviewing on research on hurricanes and things like this, but yeah, the main added complexity there is, so you don't have any more two layers. So the, the fire threat and the pedestrian decision movement, but you have the traffic as well. So you have an additional layer and many people even argue with that, we should even neglect , the pedestrian layer and just put a much more effort on the, there is a lot of controversy or because there are some, especially in the traffic modeling, they argue that it's pretty much equivalent. If you just put the pre evacuation response curve to a traffic model, uh without, uh, modeling like the, the pedestrian actions

Wojciech Wegrzynski: 51:38

So the time for for person to enter a car and start evacuating properly

Enrico Ronchi: 51:42

Exactly. So that by basically ignoring what happens in the decision-making stage, but just like aggregating this in a microscopic way. I'm not a big fan of that approach. I have to be honest because I still feel that there is a value, even though it's not necessarily something to do with the microscopic bottle, you know, to like, agent-based, it gets, you know, for large-scale computation very heavy. I understand that. But even if you do it in a microscopic way, I think there is a value in trying to break down the pieces of the different phases and try to understand what could have an impact rather than just putting everything together into curve and end. It's a bit like the problem that we face in pedestrian model in a different scale. So, but that's a modeling now. We use distributions because we don't really understand. Very well yet, actions could happen? Maybe we have some conceptual model or theories that explain like how people take decisions, but we don't really know in detail. And, and I mean, on the bigger scale probably the traffic models often, they don't really want to dig into what happens in the scale of the building. They just care when you enter a car and, and I think that's not the right approach, but maybe Rino I remember we talked about this a few times, right? I don't know. Maybe you have a different view You you are very close to the traffic modeling people probably even more than me.

Rino Lovreglio: 53:01

No I think in this case the main issues that, , we have is that we try to use the, those, distribution for wildfire. I think that when we use this approach for building, we don't get much in trouble Because it's a really contained space, but when we start thinking, like you can use the distribution that has been observing a completely different scenario todesign and, or check the safety one other completely different area is quite critical. And I think that we are still in the stage in the wildfires field in which we we know pretty much also here, what are the factors that affect the decision of uh, households, but we don't have yet enough modeling skills and data to do a really accurate prediction of what will help in different community. And that that's the first things that I need to bring regarding their, if we need to consider pedestrian or just traffic also in this case. It really depends on the scenario that we are focusing because we know that in some case you can see even just Google, it could just go on YouTube. You can see that some cases some component, of evacuation is driven by foot. It depends what, what are the safety, scheme in place? And sometime you need to use vehicle. It depends on the area that you're considering. It depends on the scenario that you are facing out. It depends on the scale of details that you want to have in your simulation.

Wojciech Wegrzynski: 54:34

I think it's also connected to the preparedness because if your preparedness level is high and you're like understanding that, such an event as mass evacuation must happen before the fire. And then you can plan better for like use of vehicles. But if you see footage from, I don't know, Greece or Turkey, the recent wildfires where people are in a way surprised by the fire when having the, their holiday time at the sea. And they are basically forced. There is no more time for vehicular evacuation towards the mainland. The only way you can go is towards the sea and be picked up by boats. So this is a completely different, setting in which these vehicle components, or at least the road vehicle component is cut off. And all you have is, is pedestrian movement; but , this is a response , not preparedness planning.

Rino Lovreglio: 55:20

And we, we might acknowledge that there are some community that are much more prepared to wildfire because they've been experiencing them. quite a lot. If you think a lot of Australians and American .Instead that are new country that didn't even know what was like, what a wildfire will look like, that, are starting experiencing them. Because climate changes, , is redrawing the maps of the area that are under risk, Enrico might tell you that after the first serious wildfire occur in Sweden started receiving so many calls,

Wojciech Wegrzynski: 55:55

What's

Rino Lovreglio: 55:55

info

Enrico Ronchi: 55:56

it's like, I will say that the amount of wildfire research in Sweden has been boosting in the last couple of years. I mean, a couple of years ago, it was really me and a couple of brave Fox. They were looking into the topic here in Sweden, but now it became really a hot topic because after the summer 2018, which was incredibly, hit by a series of, , wildfires here in Sweden, uh, everything changed in the perspective of people. You know, especially when we look at the. population that traditionally, they have a lot of a wildland-urban interface. So here in Sweden there is a traditional liking to live close to the nature and being really in areas exactly except so you're really potentially very much exposed here. So it became really hot topic, But the, I, I just also want to bring up going back one second to the topic of what we're should model on the, on the, in the wildfires. I mean, what we really need, for instance, if we really want to go down the path forward of a simplified microscopic approach, they use just the traffic model and some sort of distribution for everything that happens before we need to categorize what we have in a narrative. We have not done this. So how does the percentage of households versus hotels. Effect a pre-evacuation response curve of wildfire. We don't have a clue because you know, it's a completely different story. If you are in your property that you want to defend, you are more, much more attached compared to if you are in a hotel that okay. If there is a fire, can't wait to leave here. I don't care that I leave everything behind. So we don't have that research yet. The only thing that I've seen now, that has been done. Some research in Australia on this is on defining archetypes of people. So what, what different people in different condition will do? There is some work done in Australia on this I've seen recently, but still we cannot really quantify with the microscopic model. We don't have the tools nowadays to quantify with a microscopic model. What is the let's say the impact that certain number of households, or imagine if you have a hospital I mean

Wojciech Wegrzynski: 58:01

Yeah

Enrico Ronchi: 58:02

different store either you go down the route of

Wojciech Wegrzynski: 58:05

the school And if you're in the school season or not in the school season right

Enrico Ronchi: 58:08

Exactly So I mean I think that for the knowledge that we have now, we have to go down the path of combining pedestrian and traffic modeling. But you know, I, heard the different opinions on this

Wojciech Wegrzynski: 58:20

Yeah I've touched this subject. I brought it up because the episode I wanted about the future of evacuation modeling and as far as buildings are sexy, interesting, you know, and there's a lot to be done in the buildings here. It's, not a pothole to, to jump over. It's a whole canyon , of problems to solve, to, to have an integrated approach of fire and evacuation modeling outside of a building in this urban interface. We're almost approaching like a global scale model because if you just add weather to that you may have like continuous prediction system of what to

Rino Lovreglio: 58:55

on this I can even give you some heads up that we are, as Enrico was saying, we are trying to use GPS data and try to observe big chunk of population. and track them their behavior. And there is a lot of work I'm doing led by university of Florida by Dr. Xilei Zhao and with her, we , are trying to understand when people decide to move and depending on the area where they are, if they are their home, try to understand how they move in the, in this scenario. If they go directly away from the scenario, if they need to stop in intermediate place and those big data can help us to handle some of those question,

Wojciech Wegrzynski: 59:35

Hmm

Rino Lovreglio: 59:35

but it's still challenging because you get, you might think, ah, big data is the solution for everything. You, you get some benefits of big sample size, but you can. the uncertainty that you don't ask people what they are doing while you're observing there. So I think it's one of the place where if I was an funding agency, I will put much more, , research funding. If I will pick now between building and wildfire will probably, pick more about wildfire because especially because we are having this new wildfire even New Zealand, if you see the statistics of the number of wildfire that we are having here, and the number of people that are forced to evacuate every year is getting pretty different compared to what used to be 30, 40 years ago.

Wojciech Wegrzynski: 1:00:23

Okay for, last thing because I I've previously mentioned that the pre distribution times and, and distribution curves and stuff like that are very like European centric. And in a way also modeling is very like connected to the rich countries, buildings and stuff. And there's like 1 billion people living in informal settlements. Do you think, the modeling can help us tackle the issues of managing safety in these type of communities and well, obviously yes, but but to what extent, and the second part of the question is the current state of modeling sufficient to model

Rino Lovreglio: 1:00:57

can I get that

Wojciech Wegrzynski: 1:00:59

Yeah Go on

Rino Lovreglio: 1:01:00

No I can tell you the story and Massoud my friend is a prof. in the Karachi on, at the university. We'll be happy to hear that at some point it's like now, now we are trying , to write a paper on an evacuation drill that we run in, in Karachi, in Pakistan. And when we were planning the drill, I told him, okay, we set up the fire alarm and then we, we see what happened. And he look at me and say, you know, you realize that we don't even have a fire alarm in this library.

Wojciech Wegrzynski: 1:01:28

okay

Rino Lovreglio: 1:01:29

That's the issues. So We have some countries in which the, unfortunately, the safety issues are not the priority yet because in this country, for instance, they are still coping to make sure that the building won't collapse like a pancake in some case, and killed most of the people. So they need to invest much more money on focusing on other safety

Wojciech Wegrzynski: 1:01:51

Yeah but that's why I think it's maybe not an obligation, but it's something that we should like try and do for them because they have a bigger problems on their heads to, to worry about the emergency response, even though it's serious probably the access to sanitary facilities or the structural integrity of the, of the whole thing is it's more urgent for them. So I would be surprised to to see a local researchers developing solutions for this problems where they have something bigger on their mind

Rino Lovreglio: 1:02:23

I'm personally try to help with this kind of collaboration. I mean from my egoistic point of view can say no, and I just work with the specific university, but now I find it it's, it's even interesting because you might discover running drills in this. For me, it's a completely different scenario to discover something that we never observe in a let's call them Western country. And for me, from a scientific point of view, is exciting to discover because if you even think I've been in Karachi, I've been in Pakistan. And for me, it was like a cultural experience that it blows my mind because you can see if you think that Italian traffic is crazy, go to Pakistan or India. Yeah. And for me, yeah, I was just with my smartphone recording the traffic, because for me it was something unbelievable, but saying that even the dressing code that they use there is completely different from our blue jeans in which you have was a different way. to Walk and perform in case of a

Wojciech Wegrzynski: 1:03:24

of changes the velocity speeds and stuff

Rino Lovreglio: 1:03:26

and So far there is only one paper that was published about Saudi Arabia addressing code impact on evacuation. So there is a lot of exciting research that can be done and to discover.

Wojciech Wegrzynski: 1:03:39

Enrico, what about modeling? Because when you design a building, you have the layout. So that's pretty much the building that's going to be built when in informal, it's more like let's say or organically a built environment Right

Enrico Ronchi: 1:03:53

Yeah but I as I say this is it's not technically impossible to represent. I've actually seen as there is some pilot work that has been doing now on trying to represent the informal settlers. Basically some people, I saw you some sort of machine learning algorithm by taking satellite photos and making it into some sort of a layout. I it's still work in progress. Too much, let's say yet in terms of research, but there is a few groups around the world that are starting addressing the problem. I know. I mean the general problem with formal settlements, we have seen the, now, even in South Africa, I've started looking into this research and they have a research group. They're doing fire science,

Wojciech Wegrzynski: 1:04:36

Stallenbosch, yeah a fantastic group

Enrico Ronchi: 1:04:37

Exactly, with Edinburgh they've been working. I know there are some companies that popped up that

Wojciech Wegrzynski: 1:04:42

Yeah, Kindling by Danielle Antonellis is that's fantastic effort

Enrico Ronchi: 1:04:47

Exactly So I think these are all valuable efforts. But again, in the big picture, we, we as researcher, we, we, we sometimes hit a wall when we try to get funding for those kinds of activities. Because I mean, there are some funding agents that are more sensitive to this type of humanitarian work, but there are some very, let's say engineering minded, the Western world minded, the too Western world minded focused capitalist focus that it's very hard to convince them, to give me funding to study what's happening in Bangladesh or in Pakistan

Wojciech Wegrzynski: 1:05:19

we were actually trying to get the grant for wind caused damage on informal settlements, like three or four years ago, maybe even five years ago with, with South African people. And we didn't get the funding. The comments was like, we're, we're not sponsoring that it was legit. Like we're not interested. And then two years ago, something has shifted because the Iris project was awarded to Edinburgh where they focused ultimately on the fire safety of this. So it's also maybe a shift on politics that the changes. So you have to be at the right place at the right time to get this funding which is in a way ridiculous

Enrico Ronchi: 1:05:54

And I mean going back also to, the topic of wildfires, I think this is totally applicable also there, because the level of preparedness and the level of let's say awareness even of the problem, that there is changes, varies widely, depending on where you are. and often we have the Western world in mind, but, you know, wildfires are not a problem only on the Western world so, and that's why I've always been pushed for trying to develop tools that are free and open. You probably know that together with, uh, with a group of people, we are, we are trying to make an effort to develop a tool called the WUInity, that its a and open for everyone. It's with the Imperial college, with Guillermo Rein and a couple of other folks in RC and NFPA so. Trying to put something out that could be used by anyone so that even if you don't have huge resources, you still have something to use because otherwise, you know, the risk is that if you put, if you develop something that is a commercial too, I mean, who's going to use those tools is going to use, who can afford them. And that's a bit like I understand it that if you are a company, you have to pay your bills and so on. But as us universities, we probably should have slightly different mission

Wojciech Wegrzynski: 1:07:03

And now as companies start to grow around this topic and gather that they may be our most successful, you know, in obtaining funds,because they, they can go with a slightly different agenda than you can go. with your projects, your usually time limited time constraint, the deadline is next Thursday. And either you get a partner or not. So that's, that's kind of limiting. Okay guys. Thank you so much. , I'm really thankful for it, for this conversation. It's, it's been very interesting and we've touched so many aspects of modeling and I have like ideas for seven more episodes now So thanks Thanks a lot for for coming here Uh

Enrico Ronchi: 1:07:42

thanks a lot Wojciech It was really nice to talk to you. And as I said, it's the beginning of many conversations on this topic

Rino Lovreglio: 1:07:50

I'm pretty sure that my fiance will be happy for a new podcast because whenever we have this, I'm the one washing the dishes.

Wojciech Wegrzynski: 1:08:00

Yes The sounds restrictions of the, of the podcast recording , made Rino

Rino Lovreglio: 1:08:07

yeah Now it's going to be my pot-cast

Wojciech Wegrzynski: 1:08:09

we can arrange that even daily if you want Well

Rino Lovreglio: 1:08:12

probably next time Let's do it morning time for me And

Wojciech Wegrzynski: 1:08:16

What do you mean

Rino Lovreglio: 1:08:18

for uh for the poor guys on the other side of the work and

Wojciech Wegrzynski: 1:08:21

Okay The challenges of, of, a global village.

Rino Lovreglio: 1:08:25

thank you for having me

Wojciech Wegrzynski: 1:08:27

See you around And that's it. That was another action packed podcast episode, full of really intriguing thoughts and very vibrant discussion on the state of evacuation modeling. I hope you enjoyed that and you've picked some things that are useful for your everyday practice. For myself, some conceirns were really, interesting to hear them directly from the experts in the field, like the needs to replace the fundamental diagrams with something even more fundamental. So in the way that you're supplying the data on what your population is in terms of maybe gender, age, backgrounds, and then models should figure out what to do with that, that time to, to assess the proper pre-evacuation distribution, proper, speed versus density, equations and stuff like that. That's, it's really intriguing because now we are supplying that data directly using one or two relations that are written in the SFPE Handbook. And maybe that's not the best way, probably the easiest, but maybe not. The best one to, uh, to use in, in modeling. And the second really interesting thing was when, and Enrico mentioned two way coupling between CFD and evacuation modeling. I must say I've never thought about it in this way that we should probably include in our CFD analysts. The effects people have on the, on the state of the building at the moment, in terms of opening doors, windows, moving around, playing with systems. If you think about the most simple pressurization system, the state of its operation will always be dependent on the state of the doors in the state of the doors will be directly connected to if someone is using them at the moment or not. And that's a really, really powerful thought. Because, when we do analyze the safety of our buildings, we often run this simplistic, calculations where, actually the calculations are super advanced, but the scenarios are the most simple we can get. We, we simplify everything. We assume that doors are closed, that no one opens or fiddles with them. We assume everything works like it should. While in reality, that might not be the case. And you might end up with a completely different scenario, completely different operation of your building and in consequence, completely different operation of your systems. So if we could include the human behavior in this equation in CFD modeling, that actually would be phenomenal and that could open a whole new world of analysis in which you could truly test if the systems are fail safe, if there is a single point of point of failure for a system that makes it useless. I think that that's like really powerful thought. It's, it's really resonating with me. And because we are preparing a huge round of full-scale corridor, smoke control system, experiments that I'm going to tell you more in the near future. It's something we need to include in our experiments and it's going to be fantastic. So I am really thankful for Enrico to poaching that idea to me. A lot of credit goes to him, if not all the credit. So yeah, that's, that's something that that's already like blowing up in my mind and, and it will change the way how we approach this modeling and experiments. And maybe this is something that, that started a new, better future for, for this field of fire science. I hope it is probably im exagerrating, but I would really like it. So tell me what you took out of this episode. Tell me what you've learned about evacuation modeling and the future of it. Maybe you see a different path for the evacuation molding in the future. Maybe you have some ideas how to make evacuation modeling better. Please share them with me and with Enrico and with Ruggiero and we will very, very appreciate that. And we'll try to respond to you and, uh, maybe just like Enrico did, telling me that a simple idea. Maybe you, you will also change someone's whole research plan for the better and, uh, yeah. As a final announcement. Um, I would like to say that I was very pleased with how this episode went. And no Italians were harmed by sending them pineapple pizza. They are very safe. Their pizzas are like, they're supposed to be without the pineapple on it. And I hope they're very happy. So yeah. Thank you to my Italian guests. This was a fantastic chat with you. I had the great fun. I've learned a lot. Yeah. The ideas are booming in my head so what else do what else to expect from, um, from a great interview? Thank you so much. And to the listeners. Thank you very much for being here. I appreciate you tuning into the Fire Science Show every week. I appreciate you sharing that on your socials with friends, with colleagues in their offices. I really hope you keep doing that. I really hope that podcast keeps growing because. I am getting more and more ideas for new episodes for, for new things I can do for you to serve the audience better. And yeah, let's, let's create together, a better fire safe world. So thanks for listening. And that's usually see you next Wednesday. Bye.