125 - Enhancing Fire Safety Through Data: EU FireStat Project with Martina Manes and Mohamad El Houssami

The Need for Better Fire Statistics

Speaker 1 0:00

Hello everybody , welcome to the Fire Science Show . In the 120-ish interviews I've done with people , I've never heard someone go like we have a ton of excellent quality statistical data about fires , we just don't know what to do with it . It's usually quite the opposite in fire science and engineering we don't have great data and we really do need to figure out what to do based on this data . And it is quite obvious we always need better data . I mean , how many times great podcast guests in this show have told us that having high quality statistical data that covers fire incidents is the bottleneck in the growth of fire science and engineering ? And it's not only seen by the fire professionals . I'm really glad that the issue was also seen in the European parliaments , where MEPs decided that we need a pan-European look on the fire safety statistics . So this ended up being given to DigiGrow , the directorate of the European Union , which you should recall from the previous podcast episode about mapping the state of performance based engineering . A great project was created , a consortium out of multiple institutions led by AFFECTIES , called the EU Fires Stat Project , and today I'm hosting two people from the project Mohamed El Husami from AFFECTIES , who was the leader of the project , and Dr Martin Damanas from the University of Liverpool , who is the first author of the papers that summarized the results of the project . And Mohamed Martin will tell me not just about how we should collect data , but also answering important questions on why we need better data , why we need it at European level , why harmonization is beneficial for everyone and what can be possible once we start to collect data and , in some cases , to start to collect good data is not such a huge effort after all . To fire safety engineers , this may not be directly important to your everyday work , but , trust me , having great statistics is fundamental for our discipline , and I think understanding the background on how statistical data is collected today and how it can be collected in the future will help us all first , guide the legislative decisions happening in our countries , but also help us take better decisions that we base on the statistics that are available to us . So I highly encourage you to listen to this interesting episode about the EU FireStat project . Let's spin the intro and jump into the episode . Welcome to the Firesize Show . My name is Vojci Wynzinski and I will be your host .

Speaker 1 3:09

This podcast is brought to you in collaboration with OFR Consultants , a multi-award winning independent consultancy dedicated to addressing fire safety challenges . Ofr is the UK's leading fire risk consultancy . Its globally established team has developed a reputation for preeminent fire engineering expertise , with colleagues working across the world to help protect people , property and environment . Ofr is calling all graduates , as it is opening the graduate application scheme for another year , inviting prospective colleagues to join their team from September 2024 . If you take this opportunity , you'll be provided with fantastical practical immersion into fire engineering and a unique opportunity to work with leading technical experts in the field , while learning the skills critical to becoming a trust and consultant to clients . If you would like to check out this opportunity , please visit OFRConsultantscom for fertile details and instructions on how to apply .

Speaker 1 4:05

And now back to the EU FireStat project . Hello everybody . I'm today with people behind the EU FireStat project Dr Mohamed El-Husami from Effectives hello , hello . And Dr Martina Manes from University of Liverpool hello , hello . Such a nice representation of this giant , giant project . And the EU FireStat project was about closing data gaps and paving way for pan-European efforts . My first question is do we really need better stats ? Actually , that's a very non-honest question . Of course we need better stats . One is saying that . Perhaps a better question do we need better statistics at European level . What was your thinking when this project was born ?

Speaker 2 4:50

Today we're used to talk about statistics , fire statistics and all of that .

Speaker 2 4:54

But when you compare to what we have as statistics in other fields , we really don't have much . So if you look at the variables that we can collect , there's maybe a number of fires , there's a fire death , a number of injuries , and if you're in a very sophisticated country you might have some more variables . And still , when we talk about these numbers for each country , it means something different in each country . So the way it's collected , the meaning of each word is different , which means that the numbers are different and it's very difficult to compare from one country to another . It's things very , very difficult . So we have numbers , let's say , and some figures around fires , but it's an overreach .

Speaker 1 5:42

Martina , how did the environment look when you joined the project ?

Speaker 3 5:45

So , first of all , my research background was already in the investigation of our statistical data and I worked quite extensively in the data published by the Home Office in England and also in the US by the National Fire Administration based on the NPRST . So my background was already in the first statistical analysis in terms of evaluation of real performances of building when they are affected by fire , and the reason behind the investigation of fire incident is also linked to a better understanding of the real performances of building and when they are affected by fire and also the effectiveness and optimization of fire safety measures as well . So it was in this slide that the investigation of fire statistics is again relevant . It also represents the fire scenario , the likelihood and consequences of fire scenarios , because usually when we have experimental campaigns , the limitations are usually due to the constant scale and the possibility of testing those fire scenarios . So we could end up with that in endless what if ? Question .

Speaker 3 6:56

And this is where the first statistic actually has anything in control in terms of effectiveness or safety measures or performances , quantifying not only the likelihood of a fire in different property types so residential , non-residential but also the consequences as well in terms of fire spread , damage , fatalities and casualties as well . So this is the slide where I joined the EU Firestrike Project , and the EU Firestrike Project was also an improvement in my research as well , because the aim was to , first of all , to map the existing fire statistics in the 27 new member states , but we also investigated eight other countries , including the UK , the US , new Zealand , australia , russia , norway and in Canada as well , and so there was also an helpful and very insightful approach in order to determine which data were available , what was included in the data and which differences were present when comparing the data .

Harmonizing Fire Statistics for Better Data

Speaker 1 8:02

In the project it said that better data leads to reduction of death and injury . I guess it's not just the data per se , but where this leads . Can you comment on how actually having better data can lead to a safer fire environment ?

Speaker 2 8:19

Yeah , so to understand where this is coming from . So the correlation , of course , is not that direct . There's quite a big shortcut . But to know what's coming from is . This project was born because many people from many countries have been asking to kind of harmonize data . Because every time you want to push for a policy on fire safety , well , people ask well , can you give us statistics on the numbers of fires ? Why do we need to implement this policy ? Or if this policy or if this fire safety measure has been done in other countries , well , what was the outcome of it , etc . And we don't have those numbers to justify how well these fire safety measures are actually saving people or not . So people from different countries have been asking for this for years and there was work done at the European Parliament pushing for this and the European Parliament actually decided to fund this pilot project and then gave it to DGGROW to commission this project .

Speaker 2 9:29

And then there was our consortium that was built . So this is where it comes from , and the idea is that behind this , once we have statistics that are comparable from country to another , the idea is not to see which country is best or at least , but the idea is to be able to share practices and try to see if we see trends , or if you implement a new policy , are you going to have improvement , or your fire situation or not , etc . So this is the idea .

Speaker 1 10:01

Martin , I said the first step was to map the existing fire statistics and the terminology is a part of that . Can you give me some examples how the terminology could vary through the Europe and what issues that causes ?

Speaker 3 10:15

Yes , so when we map the first statistics in the 27 human states and the other countries I have also to make an introduction so we choose the eight other countries based on their complex and structure for statistical collection . So they were choosing based on the consideration that they had quite extensive experience with fire statistical collection and they also had quite complete data sets and we are mapped the first statistical variables available . We noticed that mainly the variables were focused on the description of the fire incident , so the time , date , the location and the property type , and then also to the variables related to the life safety , so fatalities and casualties , and only those countries where we had an extended first statistical collection . Then we could find also variables related to the item , personnel , the room of origin , the area of origin , the presence or absence of alarms or automatic extinguishing systems and several other variables . That provided us with a comprehensive description of the fire scenario .

Speaker 3 11:32

When we talk about the terminology , just to give you some example , if we talk about fire incident , in some cases , in some countries this is referred only to the actual number of fire incident attended by the fire brigade .

Speaker 3 11:45

In other countries this is referred to fire incident . The data for fire incident includes fire incident and the close explosions , and in other countries as well , fire incident , explosion or false alarm is also included . When we talk about fatalities , in some countries these are just referred to the numbers of fatalities or injuries coded at the fire scene . In other countries , these values and this record are then double checked with the values and the data provided by the medical institute After a period of time that can go from some months to six months to up to a year . So you have quite some differences , and this is also what supported also what Mo was saying before , that when we then compare the data , even if the terminology is the same , but the data could actually be referred to other things , and the differences are not just in the terminology but also in the data collection methodology and also in the quality assurance process applied to the data as well . So there are several differences that we noticed when we were mapping the existing fire statistics .

Speaker 1 13:00

For the most basic definitions , like what a fire is , what's a fatality , what's an injury . Have you have to come up with some harmonized definition that could be implemented everywhere ?

Speaker 2 13:09

Yeah , exactly this is what we try to do . So , to add some more examples , I live in France , so I'm more familiar with the situation in France . Here In France , what you find in statistics about fire death basically counts people that die on the fire scene . So if someone is still breathing , injured and transported to the hospital and then dies on their way to the hospital , or even in the hospital or a few days after , is not considered as a fire death . And so there was a work done a couple of years ago to try to see well , how much are we missing ? If we count people who die after the fire , there's about 30% that we're missing from the number .

Speaker 2 13:54

So this is enormous , where in other countries , like in Sweden , they consider these people that die after because there's a work on people that go and look at the medical record and then they include them in the fire statistics . So it's not easy to do but it's feasible . And for fire injuries , again , what is the threshold ? When do you consider is someone injured ? I mean , if you break your nail , is this an injury ? Or if you just don't feel well , is this an injury ? So if you look at some statistics , you will see France and Italy . They have about same number , let's say population , and the same numbers of fire per 100,000 inhabitants . However , there's like a factor of 10 between the injuries . So do French people get injured more ? No , of course , there's the definition that is different . So we digged into that . Some countries actually have definitions , some countries , other countries , well , they don't have anything written , they just for them . It's common knowledge , you know like , oh yeah , it's an injury .

Speaker 1 15:03

I mean , were you interested in the numbers themselves , like , were you comparing the numbers or for you it was just methodology and how their data is acquired ? Or were you also looking into the numbers ? If the numbers are actually between the countries are similar .

Speaker 2 15:18

In the beginning we looked a bit about the numbers to see if you have similarities , if it's completely like something really off , something off or not . But the goal is really to look at the methodology , the definitions and all of that . So , which is why , if you go on the website and download the reports , you will see in the first reports we show some numbers to compare them . After that it's purely on the methodology .

Speaker 1 15:45

Okay , cool , all the links will be in the show notes , so if anyone listening would like to go through the reports , if you're one click away from them , they'll publicly available , following your way of thinking . You said that between the countries there would be discrepancies like who's what's the fatality ? Is the person at the scene , is the person at the hospital ? But firefighter , for example , has the knowledge about the scene . It's also about who and when and where collects the statistics and at which moment you define . Okay , the fire statistics for this fire are complete , so I assume the process of collection was also a big thing , right ?

Speaker 2 16:23

Exactly Martina . Maybe you want to Ante .

Speaker 3 16:25

Yes , so even the collection is very different from one country to the other . So generally the firefighters they collect information when they return to the fire station after the extinguishment operation and the evacuation operations as well . They return to the fire station and they fill in a form . These forms can have different names , so , for example , in England it's called the incident recording system , but of course every country has its own form .

Speaker 3 16:58

But , however , the fire and rescue services are not the only authority collecting the data . So in some countries the data are collected only by the fire brigades on a voluntary base . In other countries there could be , for example , fire investigators after the fire attending the fire scene . We could have police collecting the data . Of course we would have also the insurance companies based on insurance claims and , as I said before , sometimes the data , especially the data related to fatalities and injuries in some countries , can be then also provided by the medical institutes . The variation and the differences are not in the collection , and not only in the collection authorities , but also how the data are collected at different geographical level .

Speaker 3 17:49

So we can have a collection at national level or state level . We can have a collection at local level or federal level . And then , of course , even how the data are done , the quality assurance process applied to the data . There could be a first quality assurance process applied by the fire brigade , so at local level , and then this is also double check at national level , or the quality is just ensured at the local level . So there is a big and wide variety of collection methodologies as well . And when we also discuss about how can we access data , so in some countries there are publicly available data sets , in other countries they are published in a form of an annual report and in other countries then the data are just private data and you cannot access the data with the confidential agreement . So you can see that the differences and the scenario in terms of international first-class cases can be very different , not only for the terminology but also in all the other aspects as well .

Speaker 1 19:00

I also assume if the sources are not in agreement , like if firefighters have different statistics from insurers and police and medical have their own and they all contradict each other that creates a very difficult , low-making situation in that country .

Speaker 2 19:14

Exactly .

Speaker 2 19:14

I mean when you look at other data , when you're talking about demographic data and things like that they all usually come from statistical institutes where they have been controlled with quality control measures and real statistical tools , this type of data within professionals about this field . However , for fire statistics it's very different . Usually , as Martina said , most countries it's collected by firefighters and the reason behind that is because firefighters they would publish annual report that shows their performance and they don't only collect this type of data , they collect also maybe the number of fire trucks they have , the numbers of personnel , fire trucks , equipment and stuff like that that they use , how geographically they are , number of intervention . All of that because it helps them usually have to justify their spendings and maybe to ask for more and things like that . So it's understandable . However , for us fire engineers , we want the statistics for something else and also for policymakers . They might need them for even doing something else .

Speaker 2 20:26

So this is where the problems start and , as you said , insurance companies . Well , they have owned fire investigations investigators . They will come and try to determine maybe the cause of fire or the economical laws and things like that . You also have , for one fire , you can have multiple insurance claims . This will give you multiple fires . You have fire that's spread in a building to different apartments . You have multiple claims and multiple insurance companies , so the work to try to gather all of this together is really , really complicated . Then you have the police . That are mainly , if they're doing an investigation is to know if it's maybe intentional fire or not , or to see why someone died . So the purpose of the statistic is completely different , which make the work difficult to harmonize all of this .

Speaker 1 21:26

You also seem to be asking quite a challenging task to whoever is gathering the statistics on the side , because as long as fatalities or even injuries , you can probably define that and let's say that's the easy one , if any of those is easy . But you also ask about quite detailed information about building the original fire , the size of the fire , the way , how to estimate that . It's brilliant things . If I'm preparing a grant which I actually am right now and I had beautiful statistics of how many fires of this size occur and how many of these size , it's a fantastic introduction to any funding project paper . But it's perhaps challenging because those people won't have time , it's an added work to them and that's quite a lot of paper to fill Exactly

Collecting and Analyzing Fire Incident Data

Speaker 1 22:16

. Was this considered by you ? It's obvious that you cannot just put everything like give me auto-cut drawings of the building and mark where the fire started and where it went right . You cannot go that far . So how did you balance it out ?

Speaker 2 22:28

This is a very tough decision to make . Well , this was a bit anticipated because it was one of the tasks of the project . So what happened is that on board with us , we had an FPA part of the consortium and they clearly told us from the beginning watch out , there's a big problem in the US . We tend to collect a lot of data , which means that firefighters , after they finish their work and at night they are filling the data report . They have to fill a lot of numbers and a lot of variables and basically , it's rarely done as it should be because , well , it's not their job , you know , essentially it's their job is to go and fight a fire , you know , and this is the boring part . So it's not done very well .

Speaker 2 23:16

So if we want to suggest to collect some data , well , we had to carefully choose what variables we want them to collect , and for that we started by running a survey that was sent to all the public authorities in Europe .

Speaker 2 23:34

So public authorities mean Ministry of Interior , mainly , and fire regates and some research institutes about fire and fire safety , and we analyzed this data to see the response of the survey , to see basically for them what is considered the most necessary essential data to be collected that would help them in policymaking or that would make sense , et cetera . And at the end we also tried to judge a bit OK , is this data useful , or is this easy to get or not ? And also to compare to what is already collected by most countries . Because if you start saying , oh well , I want to collect something very complicated like the cause of fire , you know this is all . You need A fire investigator for each fire . It's impossible .

Speaker 2 24:28

And so we ended up with a list of eight variables to be collected as tier one priority . So those are the numbers of fatalities , the numbers of injuries , the age of the fatalities . This is something that is usually collected by most countries , so for them there's need to be tweaking in the definition , but that's it . And then you have something , stuff like the primary causal factor , the type of building , incident location , incident time and date .

Speaker 1 25:02

Ok , once you have those eight variables , well , we already have a very good start , but this is like the bottom line zero , the basic ones you absolutely need to collect Exactly , and then on top of that you build additional ones .

Speaker 2 25:14

Yeah , additional ones . Well , it could be for the countries that already are collecting these data and it's a very easy task for them , so they're happy to collect more . Or because , well , they did the exercise and now they want to anticipate in 10 years what I should be implementing and try to build on top of that . Or because , well , they are already collecting some other data and they can harmonize them , and so we added some six more variables , like the number of floors , the area of origin , the fire safety measures that are present and things like that . Basically , this is the juicy part for us as fire safety engineers , because this is where , well , you can start to understand and learn .

Speaker 1 26:03

Martin , I'm tier three . I see some really juicy ones , like time between fire at arrival and withdrawal , a reason for failure of fire safety measures , property damage . At some point they stop being an obvious number you report , they start to be an opinion of the person who assesses . So my question to you how does one make sure that the quality of data is consistent over the repods ?

Speaker 3 26:29

Yes . So first of all , in terms of the quality of data , as we mentioned before , it is extremely important to have a clear definition not only of the far statistical variable but also definitions of the classes included in each variable . The other thing is that , of course , training is extremely important at different levels . And then one thing that in the most complex and structured far statistical collection level is that some of the variables are automatically recorded in some cases . So , for example , the time of the call is recorded , the time of arrival of the fire brigade at the fire scene is automatically recorded . So in this case , having automatically recorded variables of course reduce the errors as well , in terms of the presence and effectiveness of automatic distinguishing system or alarms , for example . In the most complex and structured far statistical collection we already have those variables . And when we talk about fire spread , just to give you an example of the classes that are included , usually we can have no fire spread limited to the item first in item , limited to the room of origin , to the floor of origin affecting two floors or affecting the whole building . So these are usually classes that can be identified quite easily .

Speaker 3 27:56

When we then talk about fire damage , then usually this is quantified according to square meters , usually if we are in Europe and this is something a bit more complex , especially in those countries where we have also a sublimation of a further classification of the damage in fire , flames , smoke and water damage and therefore you can see that the complexity is increasing .

Speaker 3 28:21

But generally in the countries that we map , the information is usually mainly confined to the description of the fire incident , the fire , fatalities or injuries , the cause , and then , when we go into the damage , usually the fire spread is recorded and then the actual quantified value for the damage , the square meters or the square feet .

Speaker 3 28:45

Then this is a variable recorded only in a very limited number of countries . The other thing that is usually applied to ensure the data quality is that at central levels there are also checks , so they actually check the data and if they find some errors or inconsistencies , then they have a follow up with the fire brigade responsible for the collection in order to verify if that particular value is inserted correctly or if this needs to be then edited based on the information . And , as we said before , there is always a tension between increasing of the physical vehicles and actually ensuring data quality and reliability , especially considering that when the firefighters return to the fire scene , return to the fire station after attending the fire scene . Fatigue is also added when they fill the report . So there is always this tension between increasing the fire statistical variable to have more information but that actually have and ensure a good quality and good reliability of the data .

Speaker 2 29:55

Also , when we talk about quality measures , I mean it doesn't mean that we want the data to be 100% true . We don't go to this level of detail . So one example is the age of fatalities . Okay , let's say the firefighter arrives on the scene and at the end , while there is like one fatality , he can look at him and if he doesn't see any ID or anything , he can just say all right , this person looks about mid 40s , can put like 45 . And then when we do the quality check or when this is crossed with the medical data , then they can correct it . But it's better to put at least maybe 45 or something like that , instead of saying I don't know . You know , because with 45 , I can do something , even if it's wrong . It should have been 47 , you know , but is this going to change anything for us ? No .

Speaker 1 30:47

I think it largely connects with what you can do with the stats later on . So if you map age or occupancy or some other like personal characteristics of the victim , you perhaps are able to map the people who are at larger risk than the rest of the population . It's an obvious use of such a data . We don't need the age of victim just to know how old are people when they die on average . There's no information in that . Perhaps there is , but not that interesting to me . But if I know that , for example , elderly people living in the rural parts of major cities with the distance to fire brigade of 10 kilometers and more are at much higher risk , that means I can act on that data . Were you running such analysis to investigate , like , what you could do with the data later on , to figure out how to create the data and how to create actually the mapping , the inquiry process ?

Speaker 2 31:43

Exactly because the idea after is that you cross this data with , maybe , demographic statistics and so you know , in a neighborhood where it's mostly elderly , well then you can see the trends and the correlations and things like that . But then we had the idea also to use some classifications like youth or elderly , but the problem . Or senior or adult , but the problem is that from a country to another it changes a lot . So senior in one country could be above 65 , in another country above 64 or something I don't know . Or youth could be something between 15 and 18 . So we didn't go into that because it was a big mess between countries and we couldn't get along . So we decided just to keep the actual age .

Speaker 3 32:33

Also something related to the variables . The variables are not related only to , for example , the human characteristics in terms of age , gender or if there was a death or if there was a fatality of there was an injury , but also to the building characteristics . So , for example , numbers of floor , property type in residential and non-residential buildings and then those two fire characteristics as well . So the atom first , united had the boom of origin , like there was a fire spread . Something that we also did during the project , we also developed during the project , is that when we provided the terminology for each of the selected variables , what we did is also that for the property type , usually we have residential and non-residential . An improvement that in the project we applied was that in the definitions for the property type we also considered the mixed use buildings , so a building where there's a presence of residential , but also a business as well .

Speaker 3 33:40

So the definition that we created were , of course , based on the existing definitions , but we also tried to add some small improvement that could give us a comprehensive view .

Speaker 1 33:52

To better understand where this all is going . So this research is funded at the EU level . You said European Commission was concerned about the statistics and here we're talking about stuff that individual countries collect . So there's perhaps a compromise in creating data that is perfect for one country , for their context , for their needs , for their situation with how fire department works . In some countries it may be mostly voluntary . In some countries it will be government , public state fire service . In some countries the fire service has a say in the building codes . In some countries they want it's different right . So perhaps countries would like to do the statistics for their own . There must be a challenge to harmonize it in European level , to give one number for the whole EU . Now I wonder in your project which need was more important to have a pan-European system that's harmonized or to just create systems for each country to work , and how they connect to each other at the EU level is secondary . Which one was the prime ? Or maybe there was a third one , I don't know .

Speaker 2 35:00

Yeah , it's the third option . No , but I mean both systems can work together . This is the idea . So this is why the European one should not be a burden , an extra burden . It should be something that is based on what is already collected , and the idea is really to try to find common ground with what is already being collected so it doesn't create extra burden , so that also countries try to converge to that with time , so that let's say , for a couple of years there's the two system going in parallel and maybe they would converge , so that maybe after 10 years everyone would be just collecting this data in addition to what they also want to collect for their national purposes . It's a bit like a reaction to fire the classification . So in the beginning you had the national classifications Each country has its own stuff and then they created the euro classes and so for a while you had both running at the same time . And now , well , the national euro classes are less , more and more disappearing , I would say fading away , and everyone is switching to a

EU First Act and Future Plans

Speaker 2 36:14

euro class .

Speaker 3 36:14

And also I agree with what Moulin just said , and we need also to mention that the outcomes and also the report were the consideration , the deliverable , the suggestions that we produced during the EU First Act originated from a very extended conversation that we had with the local and national first statistical authorities in each country . So in the very beginning of the project we established connection with the first statistical authority of each country and it was quite challenging because we also worked at the project was developed during the pandemic , at the beginning of the pandemic . So everything was developed online and what was surprising was the real effort and participation that we found from the local and national authorities . They were really open and helpful and supportive to us and to the project , and so the consideration and the suggestions provided by the project are based on the discussion , that extended discussion , that we had with the local and national authorities .

Speaker 2 37:25

So now what I wanted to add is a great point for Martina , as first , thanks to the European Commission , because they opened the door for us . We had this very nice letter of the European Commission saying please help these guys try to solve the problem . So it helps a lot to have the proper people from the ministries to answer you . But then , once they started listening to us , they saw that oh OK , our project is interesting , You're actually hearing what we have to say , we were taking notes and you are making us participate to the steering committee , et cetera . So everyone was almost on board .

Speaker 1 38:04

What percentage of countries that you approached told you ah , thank God , because our statistics are horrible .

Speaker 2 38:11

Let's say we approached the 27 EU countries and we had feedback from the 27 . So this is already very successful . That's big successful . Yeah , ok , sometimes the feedback is we don't have anything .

Speaker 1 38:26

Well , that's the easiest one to solve . Just build it based on all your recommendations , actually .

Speaker 2 38:30

Yeah , yeah , yeah , exactly the other feedback is we have data but we cannot share . This was a tricky one .

Speaker 1 38:38

OK , that's a bad one .

Speaker 2 38:39

Most other countries is well , we have stuff , look at it and then we're happy to answer any questions . They don't usually say our data is bad . They usually actually say we have the best data in the world , baby , ok . But also it was very tricky because , of course , you had to overcome the language barrier and for that we were lucky , because the consortium was people from so many international people that spoke many languages , and we were always able to find a way to talk to the different languages and to cover all of that . So I want to thank everyone . Great team effort .

Speaker 3 39:22

I also think that we need to thank all the consortium members as well , because they really for their professionalism and also for the time and dedication that they had for the project and during the project , and also reaching the 27 new member states , the authorities in the 27 new member states and the other countries was a big task for each of them , so I think that we need also to acknowledge I'll do it right away .

Speaker 1 39:50

So the consortium was Effectis , a consortium leader , and then BAM , cfs-ctif , the Center for Fire Statistics of CDIF , danish Institute of Fire and Security Technology , dbi , lund University , an FPA School of Engineering at the University of Edinburgh , dutch Burns Foundation , the European Fire Safety Alliance and FAUFDB from Germany . So very broad representation indeed of European institutions Like the Firefighters . The Burns Foundation is firefighters having academics , like you , martina , having researchers , having testing institutions that's really nice consortium .

Speaker 2 40:30

Yeah , so there was an FPA as well giving us their expertise , even if it's outside of Europe , but they had so much experience in this topic . And also , to mention from Lund , unofficially I would say , there was MSB , which is the . I always forget what it stands for , but it's basically the people who deal with fire statistics in Sweden and they were very , very active in giving us back an input .

Speaker 1 40:59

So you had this Pan-European partnership to do this . You had the iron letter from the European Commission . These guys are doing it . It ended one year ago . Any real implementations effects success stories already . How is it look ? I know it's a difficult question to answer , but how do you feel about implementation of this ?

Speaker 3 41:22

Yes , so , based on the project , we have , of course , a website where people can download all the reports that we have produced during the project . After the project , we also published two papers in the Fair Technology Journal . We were invited by the European Commission during the Fire Information Exchange platform to present the work . We were interviewed by the NFPA for the NFPA magazine and also quite recently we were awarded with the UK SFPA Research Award of .

Speaker 1 41:57

UNRAP Congratulations .

Speaker 3 41:58

So the project had several outcomes and still nowadays , after almost one year after the project , the project still has a big impact on the first safety community , but also in the countries that we in the 27 year member states . So one implementation , direct implementation , is that in Italy they revised and reviewed their current practice , first statistical practice . They have implemented some of the variables that we suggested , based on the outcomes of the project , and they will also apply a new and improved first statistical collection starting from the beginning of 2024 . So there is already , there are already some countries that are actively considering the outcomes and trying to implement the suggestions and the consideration produced during the project in their current practice . So there are ongoing discussion with several UN member states at the moment and the project has quite a strong impact not only at European level but also internationally as well .

Speaker 2 43:07

So we had to add to that . At the end of the project we proposed , we also sent a survey to the different countries where we asked them well , so this is what we proposed . Do you like it ? And are we willing , maybe in the future to work on a follow-up project where we look at maybe the feasibility of implementing all of that , because what we propose is might be nice , but what's your in the field ? You might discover things that don't work , or well , you collect them differently or interpret them differently , et cetera .

Speaker 2 43:41

So we propose that maybe there should be a second step where we select a few countries or maybe not countries , maybe just municipalities or fire brigades where we try to implement this for a while , see what it gives , and then we tweak a bit the proposal and out of the survey , out of the 27 countries , 19 of them answered the survey and the 19 who answered said yes , we would love to do that , we would love to participate . Ok , that's solid feedback . Yeah , exactly , exactly . So now we're hoping that there should , there would be a second step to do all of this .

Speaker 1 44:18

And any feedback from those eight outside of EU countries that were also surveyed , or there was no feedback , will build those . Martin , you look like living with one of those .

Speaker 3 44:29

Yeah , exactly . So usually the other eight other countries has already quite structured and complex for a statistical collection . So some of the suggestions that we made during the project that we already have implemented in other countries from many years . So something that could be done in the future is to also contact them in order to understand and gather their feedback and also highlight some possible limitations or challenges or something that one is going very well so benefits as well , and so I think that they will still be part of the discussion , especially in the implementation , due to their quite extended experience with the first statistical data sets .

Speaker 2 45:12

Yeah . So we're still getting feedback , also from some countries asking them from time to time . So , hey , any news ? Do you have any ? You know we want to implement this . Can we start , or is there going to be harmonized work more , et cetera ? So it's nice to see that there is still this momentum and to keep it . We are planning to work on ISO level at ISO TC-92 . There's a working group on fire statistics to see how this work will be taken and if we can create standards based on that and to take the work outside of Europe to more internationally , to improve it also and to give the tools for the countries .

Speaker 1 45:57

Are you still in touch with the FPA ? I know an FPA is doing a lot in the space of statistics , so perhaps these people would also be interested in it .

Speaker 2 46:05

Oh yeah , of course .

Speaker 1 46:07

So for the final questions , as I'm writing this grant , so where I can find all the European statistics in one convenient place so I can just put them in my grant now I know it's not yet on the table , sorry for a curveball . Will we have such a resource that we could like we have for accidents , for example , or statistics on how many vehicles are in Europe or how many electric vehicles are in Europe ? No , there are these statistics that you have access to and this is a statistic I don't have access to . Is there any vision , like a bigger vision , behind the project to create such a resource base ?

Speaker 2 46:41

This would be the goal really . This is the end goal is to have something like you have for road vehicle . As you mentioned , we actually even looked at how they structured this so that we try to mimic and see the difficulties . They were part of the steering committee it was a DG move , I think , and we took their feedback really to include it . Today , if you want to look for fire statistics I mean there's a report , of course , but you can also go on the CTIF website . They do very good work into contacting each member state and collecting data from them and they put them in one big table , but then there's no interpretation behind it because they know that there's no quality checks . Each country basically just publishes what they want and you don't really control what they publish .

Speaker 1 47:36

That's pre-EU firestat data .

Speaker 2 47:40

Exactly , but I mean , it's still something you know . If you need it , of course , it's better than nothing , and we appreciate their work .

Speaker 1 47:47

So what was the most surprising thing during the project ? That really , like you , thought it's going to be otherwise , but it completely surprised you .

Speaker 3 47:55

So I think that the most surprising aspect was that at the beginning , it was a big task to actually map the 2017 Rambos station at the data countries . We map a total of 35 countries worldwide and at the beginning , of course , we were not sure if we were able to contact all of them , because it was also during the pandemic , so it's a very delicate situation for the entire world . But what was surprising was the response that we received from all of the 35 countries the 27 human states and the eight other countries . They were really responsive , they were really supportive , and so that was a big price . And then the other thing is that , as I mentioned before , after one year that the project has been completed , we are still talking about this project . We are still having a big interest around this project , not just from the fire safety community , but actually from the human state and the fire statistical authorities as well . So there's a big momentum around this project and , of course , for us , this is an honour and we are really glad that this is happening at the moment .

Speaker 1 49:08

I'm not surprised . Data is gold and you are here creating a gold mine . So that's a good investment for everyone and I think everyone's interested in good data . But the scale is really something else . It's quite obvious everyone would like the data , but when you see responses from literally every single party inquire , and then the interest and everything , it must be very nice and , as a researcher , that's the impact I would like to create with my work one day . Mo , what about you ? What surprised you ?

Speaker 2 49:39

So , like Martina said , I had the same surprises . I think that the pandemic for this project was a blessing . It's disguised , because we were supposed to meet in Brussels every time and basically you know it's not going to be working as good as it's supposed to be , and then we all had to switch to meeting on team or Zoom or whatever , and then the momentum continued . So it was good .

Speaker 2 50:06

One other thing is just as a project manager . It was a big project to manage with like nine partners , the European Commission , all of that , and she had a lot of fun and because all the partners really did their work and even more than what they were supposed to do . Obviously , we'll spend more time than what we had budgeted for this project , but it's because out of passion , I think . So all members of this consortium really gave it 100% or maybe 200% . They went over and beyond what we were supposed to do so that we could accomplish our goal , and I would like to thank each one of them because I learned so much from them . I mean , I didn't know much about fire statistics before going to the starting this project .

Speaker 2 50:58

And now , well , I'm here , I am interviewed , you know , but this is nice . And so we even have one of the partners from TVI I'm going to mention . She was from a different nationality . She went to the embassy of her home country to try to reach out and find some correspondent who could answer our questions oh nice . So that's the mission .

Speaker 1 51:23

That's good , this is real dedication yeah . Doing important , impactful stuff with nice people in a nice consortium .

Speaker 2 51:32

That's what people like to do , and that's what the research likes to do and this is what the project manager wants to see .

Speaker 1 51:40

Thanks . Thank you so much for coming here and sharing and to the listeners . All the Reapers are available online , the papers and everything Everything's linked below , so if you would like to learn how this is structured and what you can get from this data , then there's a lot of reading for you , my friends . Thanks , guys .

Speaker 3 51:59

Thank you , and that's it .

Speaker 1 52:01

Thank you for listening . One funny three-vibes about statistics comes from the episode with James Quintery where he told me that the big investment into fire science in the 70s was triggered by a kind of misrepresentation of the fire severity in the data of US fire incidents , including vehicles . The amount of fatalities was exaggerated like tenfold , which triggered a governmental response that we need to fund and study these types of fires . So perhaps sometimes there's also merit to having a data of little less quality . But overall , in general , we need good statistical data to really make important choices and decisions related to fire safety by guiding our authorities towards making fire safe solutions standard in our countries . And if we don't have statistics , we don't know anything . So great appreciation to the EU Fire Stat Project . This huge consortium that has dug into the data from 35 different countries came up with a standardized form and definitions on how to call things in fires , how to collect the numbers about those things , how to exchange this information and also , to an extent , about what to do with it in the future .

Fire Science Show - Episode Wrap-Up

Speaker 1 53:20

So that is it for today's episode .

Speaker 1 53:22

Thank you very much for listening . I hope it was an inspirational episode . I don't think many of us will be able to put this directly into our fire engineering , but our fire engineering highly relates on stuff like things discussed today , so always great to see how it looks behind the curtains . I hope you enjoyed it . See you here next Wednesday . Bye , this was the Fire Science Show . Thank you for listening and see you soon .