 Thank you very much for the introduction Don. A few words on the guide for scientists for quantifying the value of structural health information for decision support. Okay, this is the slide we have just seen. This is the structure of this document. It contains the approach for the value of structural health monitoring, the description of structural health monitoring. The scientific core here in terms of the structural performance adaptation and with structural health monitoring information and its modeling and principle example. You may have recognized, we have been talking about the value of structural health monitoring or monitoring information. You may have recognized also from the previous presentations that in general we cannot really distinguish between inspections and monitoring and maybe some other kinds like non-destructive testing. So in this sense we introduced the term of structural health information. This is what we are dealing with and we will see later in the presentation that there is various kinds of information we can obtain from a structure covered. To have or to visualize broad perspective we should consider the life cycle of a structure from the planning and design phase over manufacturing construction, a large area of large time span for the operational maintenance and then the commissioning phase. In all phases there are decisions, symbolized with the decision tree here. If we zoom in then we have the system states here which may distinguish in a way of risk analysis into hazards, constituent damage and failure states and then system and damage and failure states but we should also not forget the intact system. And then all these decisions are about the maximization of the expected utility. This is the foundations we are working with going back to utility theory and they are somehow related to the economical efficiency, safety and environmentally friendliness of our systems. They are subjected to some boundaries like in terms of utilities. This is formulated here as an acceptable utility and acceptable minimum utility. And yes we are optimizing across sets of decisions. From this rather broad perspective we have also been seeing today this is defeated basically from the theoretical framework. Working group one we see many exemplary decision scenarios which are part of this perspective in the phase of the operation of an infrastructure system. Yes many many this is for the state now relatively good explored. We have code and standard calibration structure prototype development and design by testing and this is what we have not heard. Yet we also have the structural health information system development. And this is also something we need to use our framework for because this is providing the utilization scenarios and also the perspective on where the development of structural health information systems should get us on maximizing the value of information basically not necessarily the precision of the information. Yes well there is a probabilistic part in this document I will try to keep this short. So this is a zoom into the objective functions for the value of SHI. This is formulated in the way of a preposterior decision analyzers. We have seen this before with combining the preposterior branch here with the prior decision analyzers. So that's the U1 which is quantified here the U0 quantified here and then our optimal decision is about the identification of the optimal information strategy the optimal set of actions depending on outcome and also the optimal action without the SHI without the structural health information. Subjected to acceptability boundaries. So this is what we do when we do value of information analysis in the sense of that we are optimizing the information requirement before implementation. We can also quantify the posterior value of SHI. So this is the situation where we all already have measurements and an outcome and then the decision is basically about here quantifying the optimal actions with and without information. This is symbolized here so we don't have a decision of identifying the most optimal strategy for information requirement. We have already obtained the information so the path is drawn here and the decision note here is gone and also this decision note here is also gone. So this stands for the decision whether to implement SHM or to acquire any information at all or not. So it may be that the value of the information may be negative and in this sense there should not any system be implemented. So this is the description of the framework for the quantification for the value of structural health information in a wide perspective and then with examples including a general probabilistic formulation. We now go into the structural health information so this part is inspired by the work of Working Group 2 what we have heard today already and Working Group 1. We have seen this terminology about information conditions we need to account for or need to be at least aware of what information we can obtain or not. So this gives us a definition for the structural health information which is information with relevance for the decisions influencing the infrastructure performance and utility. And specifically we need to model the type, content of information, the probabilistic properties, costs and consequences having these conditions in mind. How can we classify the types? We have had quite some discussions about this at this level with this guide. We keep it at a very general level, temporal characteristics, discrete or continuous and periods of measurements. We have spatial characteristics, are we on system level, are we on component level or constituent level or subsystem level. And then is the, do we have a direct information or do we have an indirect information so that's rather an indication information. What probabilistic characteristics do we need to account for? Close to what we have heard from Working Group 2 you may recognize it is about the measurement process. So that means the conversion of electrical, optical signals to structural properties. It is about the SHI installation and operation. So sometimes there's the issue where the sensor, where exactly is it? Yes, we don't know exactly but it must be here somehow. This can make a large difference if the measurement was very local like with strain gauges. Operational conditions need to be somehow modeled. Best is if they are somehow normalized. Human errors we have here, human errors we have also in the SHI data analyzes. So this is about statistical uncertainties and model uncertainties. Limited precision also of the data normalization algorithms and very important. And I think this was not mentioned so far. The dependencies between consecutive or multiple SHI information, that's doubling, multiple structural health information. So here we need to have quite some attention to it. The dependencies are originating from the measurement system. So the part of the measurement system which goes to the sensor, it's just maybe rather independent. But if the same amplifier for instance is used then we have already some dependencies. So these are the main probabilistic characteristics which we should find in our models. One way to model the traditional way is we are basing upon signal processing and detection theory. So basically we have a distribution of the measurement signal and the threshold. And then we can distinguish one or the other state so that originates in hypothesis testing. And again we need to account for the dependencies here. So this was an outline of the structural health information modeling. In a little more detail, the type, the probabilistic modeling, of course there must be the cost modeling. So what is the lifetime of the SHM system associated with the replacement interval? What is the cost of installation? What is the cost of investment? Okay, coming to the part number four, structural performance adaptation and structural health information modeling. So we may develop very similar approaches for different types of structural health information at a very local level, inspections or damage detection. Damage detection for systems may rely on the statistical and dynamical behavior of the structure and analyzes in regard to a reference state. Low testing is a way of information requirement and then we are coming also to monitoring information which may also be modeled in as a quality information where the limit state function is an inequality and we have a quality information because we know that some property, some structural property is equal to the value we measure. So short zoom into with a few formulas. This is what we know relatively well. That's the posterior probability of it could be a constituent state X and we obtain the vector of the information Z here. This is the posterior probability. What is not so clear and causes sometimes confusion is this is what we don't do in the pre-posterior analysis unless we do the extensive form and then multiplying with P of Z. So what we're actually doing in a pre-posterior analysis is we are after this expression here and this is for illustration how do we determine the probabilities of the system state of the constituent system state X1C here. This is a limit state function. It's basically the resistance minus the loading but the resistance dependency dependent here on the damage and we also count for the model uncertainties. Then we can also model the limit state function probability of the outcome Z1Z2 based on the probability of indication for all damage states. This is much more extensive in the guideline or in the guide and we also have a principle example where we pre-assess the value of SHI with simplified approaches. We assess and optimize before SHI system implementation. We use the posterior decision analysis or posterior value of information analysis to assess and optimize after SHI system implementation and we provide a starting set of literature for adding complexity to the principle example. Thank you for your attention. I will hand over now to Dimitris and we will take the discussion at the end of the presentations.