 So now something completely different. This is not a case study. Actually this is mainly for PhD students among you. Last week in Nantes there was a big absent conference which had a special session for PhD students which was called my thesis in 180 seconds. That's what this here stands for. So the point was to present the PhD thesis ongoing or even already finished one with one slide in three minutes. Nobody interrupts in those three minutes and when I first heard for it I applied but it looks simple to me. But when you want to do it it's not simple anymore because you need to present your thesis from the beginning so from the problem to motivation, your idea of solving the problem and what do you expect as a result. And you need to present it to audience that is not all of them are closely related to your topic so you cannot put 20 equations because nobody will understand you. So this is my presentation I've done but in France this little thing here it was timing and when it hits zero they turned off the microphone so you cannot continue and go over. No, you shouldn't tie me because when this was last week I practiced and last night we didn't practice. We did but not this so I will try to do it and I challenge all of you who are PhD students you don't need to present it but try to summarize your thesis as simple as you can but with clearly defined. Why are you doing something? What exactly are you doing? What do you want to achieve and what's your plan on how you will achieve it? So you have a problem, your idea of solution for that problem and what do you expect as a result which is not always quite simple. So I will start now without the timer but I'll try to do it like I did it and okay you can but don't show it to me you can just and the best thing about this session is that there was no questions after it so you just presented applause and next. So okay we can start. So my name is Dominik I work at the faculty of civil engineering in University of Zagreb my thesis is titled probabilistic assessment of existing road bridges using bridge weighted motion and the main motivation for this topic was two facts. First one is that in Croatia there are no official guidelines for assessment of existing bridges while on the other hand majority of road bridges are 40 or more years old designed according to all codes and they need to be reassessed in order for their use. So the idea was to define some assessment method which will use SHM and probabilistic approach and we started with bridge weighted motion measurements as an chosen SHM tool so that's a SHM tool which measures vehicle as they drive over the bridge and you get not only traffic data but also structural response of the bridge which is very important and we combine that with probabilistic approach and Bayesian updating of material properties of the bridge. All that combined we defined in a multi-level assessment method using SHM and in first part of the thesis we applied this method on 3K study bridges and the results show that there's a great increase in bridge reliability and automatically extension of the service life of those bridges but when you look from the perspective of bridge owner the question is is that cost effective and that presents second part of the thesis which is ongoing case study in this cost action so we tried to define the value of information obtained with this type of SHM and in the end is it cost effective. We are in process of doing that but some preliminary results are very good and we are hoping that the final result will show that using this plus value of information can lead to optimized bridge management and just for the conclusion official thesis is that implementation of these measurements and this approach can lead to extension of bridge service life overall reduction of maintenance costs and optimization of service life management and that's it. How much? 39 seconds left. I don't want to talk anymore. Maybe I was a bit too fast but that was kind of the point. The only difference was that it was in auditorium for 800 people but there wasn't there 800 people but it's a big stage like you are giving a TED talk so of course I was a bit more nervous and I wasn't hangovered. Yes. Okay, I will. So bridge weighted motion for those of you who are not familiar with it first without bridge only weighted motion. Weighted motion is those are sensors who are built into the pavement and they measure the vehicle as it drives over them. The information that they give you is how many axles the vehicle has weight of every axle, gross weight of the vehicle plus vehicle speed and it was way in motion as an idea started in the 70s and way in motion data was used but data from the late 80s was used for development of Eurocode official traffic load model for design of new bridges but they used only limited data small sample and that's why that Eurocode traffic model is very conservative and if you apply that model on assessment of existing bridges no bridge which was designed by older codes can be assessed as safe and bridge weighted motion as an idea is that those sensors are not put into the pavement because when you put them into the pavement there you cannot move them and you need to close the traffic. Bridge weighted motion is an idea that uses the bridge as a weighing scale so you apply sensors on the bottom side of the bridge, you calibrate them so there's only minimum interruption on traffic on the bridge or sometimes you can do calibration while the traffic is on so you don't have to close the bridge at all and it gives you, it even has more accuracy and it gives you not only traffic data but structural response of the bridge under the load which are very useful information for the analysis of the bridge state so the influence lines and dynamic characteristics and dynamic response which is very important and which can show you not only the state of the bridge but state of the pavement of the bearings, expansion joints etc so we, in this method we defined two types of bridge weighted motion monitoring short and long term short gives you only the structural data but not the dynamic characteristics and it can only take one day of measurements long, it gives you dynamic characteristics and also it gives you enough traffic data which if you use extrapolation methods or Monte Carlo simulations you can predict traffic load in some certain return period but you need at least two months of constant monitoring so along with Sebastian this February we built a decision tree where we defined three strategies one is without any monitoring which we use as reference one is with short term and one is with long term monitoring and we are in the process of finalization we are still defining some costs because there are many, many related costs not only cost of the bridge, cost of the bridge repair but many indirect costs because it depends, if you close the bridge it depends if there is a bridge two miles after it or if there is no bridge in radius of 30 miles it depends how many daily traffic is on that bridge and many other things so there are many indirect costs that we are trying to model but some preliminary results show that in many cases on existing bridges short term monitoring is always a good investment long term we didn't yet define but if the bridge is in bad state based on only on the load carrying capacity assessment it could also be profitable and that's the kind of whole idea and that value of information model is one of the case studies cost action but it was presented in Italy in Portugal we don't need to present it again that's kind of it Do you calculate its value of information? Do you consider the value this information has for assessing this particular bridge? Yes, we did Also the additional value of course that we said the bridge is part of the highway so the traffic going over that bridge would be almost the same traffic going over the next bridge Yeah, there are two types of data traffic data but also the structural response which you cannot use the same traffic data for next bridge but you cannot use dynamic response and structural response The main integrity is often from the traffic if you go to the traffic model of your horizontal mobility so the traffic is the same If you have a... In Germany they put these things they put them on bridges they put them on one and then they try to take traffic model and then they take traffic model Yeah, for example for development of traffic model that's enough, you don't need to monitor every bridge you find the best one which best suits you because it's not the same it's a viaduct bridge it's too high, you cannot reach it from the bottom so you find the bridge which suits you where the installation will minimalize the cost so yes, for traffic model that's enough We also said that people in Germany they have not done this equation analysis but they just put these on one bridge and then we don't put it on other bridges because they think that if you have it from one bridge it's no more cost-effective to put it on the next bridge so yes the additional information you might get from the funds it's maybe not once you know the traffic is not... Yeah, I know, but for example it gives you realistic influence line and that can tell you a lot of the state of the bearings of expansion joints it's kind of... it gives you the kind of data which you obtain with load testing when you open the bridge It's got a lot of influence line but I don't have an explanation It seems to me that influence line is for the... Well, you know, for example influence line for... if you are talking theoretically influence line for... simply support and beam sorry influence line for this, for bending moment in the middle of the span looks like this if this is a middle and this value is span divided by four so it means if you put force of one here, moment here is force multiplied with this in reality nothing looks like this it looks like at least like this because so when you're... when people were doing assessment they were doing some of them even doing simple 1D models and they are very conservative they use this kind of influence line and they put traffic load model on that and they said no, this bridge cannot hold the load so first step was we build 3D models and then we calibrated them with this influence line there are some uncertainty there but I think we got it because boundary conditions in reality in theory in bridge design are for example like this in reality they are nothing like this so we are trying with with yeah, there's no point loads and of course in reality bridge girder is I don't know 1 meter and then you also need to use this height to additionally reduce the the influence line plus you need to use the width of the bearing so it goes on and on but that's why we on each bridge we monitored so that it gives significant reduction of bending moment in the middle and we are using it mainly on road bridges up to 40 meters span because by I think script project was in Germany and they made a survey on 10 roads on trans-European roads and majority of bridges road bridges are simply supported or continuous with a span of up to 30 40 meters and mainly pre-stressed and concrete so we choose that type because this method is not for landmark bridges you cannot I don't know I mean you could maybe but that kind of bridge requires special assessment but this was for this small bridges which are kind of neglected at least in Croatia and they are really bad state