 But reflecting on day one, what feedback did we collect? And this is my perspective. I think very important feedback and points we received. It's not that we didn't know about it, but of course, we are very grateful for clarity. So the decision about structural health management should be in the design phase and in a wider perspective. It should be also the structural integrity management should be considered in the design phase. And this opens the map to designs which are also relying on monitoring and inspections, rather than being very conservative. So first point, I think the decisions should be taking the design phase, this is very important. We focus on criticality. This can also come up here. I think this is very important. It is in the design, the criticality of the components, either by design checks or in our structural reliability analysis. And I think a good example of the two good part presentations, one of which was mine focusing on the operation phase, but the next step clearly is presented. The next step is also to include the design phase here. And we need to have clarity and optimality what and when to measure with what technology. And I think here, this state in my mind, we need to address the very basics. And this is the type of the information. And here the example is the visual bridge and the damage mechanisms which have been analyzed. And then there is an identification of the centers which are able to know about these damage scenarios. And I also understood that this has been taken up in other countries. I think it was normally right that these damage scenarios have been gone through with models. And that we identified the centers that we have to write type of the information. And of course, we would like to, so it's very important to address this. This is a very good visit for our information analysis. We need to type the precision and the costs to then assign the value of this strategy based on these scenarios we are considering. So then I think extremely important is following the measurements by the designers. This is for me an extremely important one. It's a hope field in research and in engineering as I understood. What is missing here, where all is missing, I also understand that the dinars should follow the coordinates. But that's sort of work on interpreting the measurements and to relate them to the structural models. This is not straightforward. This is an effort and it requires, no doubt it requires an excellent structural engineering with structural engineers who not just know their homes but know reality. And it requires an excellent SHM engineering. So there should not be like I sometimes in projects, be some clarity of where the sensor exactly is located and it needs to be understanding very basic for square engagement measurements. This is very, very local measurements. And if you have a structural design, a simplified model and a beam model, then you can almost be sure that, so lately by mechanics, you cannot directly relate the strains you're having from the beam model to the strains you are measuring because they are much more longer. And you need a very local model. So it could be a shape model. OK, this, I think this is a challenge, but we are aware. We have identified the main structural engineering can bring benefits. It is prototyping or maybe related to prototyping. And the point here is that we would need to show the benefits of doing this. And the benefits of doing this maybe that we could think in the direction of making a new design, much more sophisticated new design and cheaper, more efficient design. So this was the point of following the measurements. They should be followed by the designers, but they are not the only ones who should be involved in this process also together with the SGM engineers and maybe also some scientists. And the extremely important point and we have been also taking up for the further development of the field of SGM is that we need our decisions now. We need the actions and we need the triggers for the actions. And then our SGMs should facilitate to save money to delay the actions because they are not necessary because we know much better the commission. I think the base here is that the clarity in the case studies is required. We have it in. It is the decision rules. So what is the outcome of our SGM strategy? And what are the outcomes? There's very different outcomes. And then base, for instance, on thresholds to trigger actions. And then the task here for the researchers is to do maybe a pre-posterior decision analysis and to go through all the possible outcomes and then to derive the decision rules which are very of information-optimum. An example for a very of information-optimum decision rule is if you do inspect and you decide about to prepare. So you could inspect. And then you can record the outcomes. So there's a detection of a crack or no detection of a crack. And then you can think of whether to optimally repair the structure depending on the outcome. So in a pre-posterior analysis, we go through all the branches. So even if we have a detection of a crack, we may repair later or we may repair just now. And we have shown, for instance, that very of information-optimum is to inspect and just to repair. So in case there is a detection. So that's rather intuitive. But it's also very of information-optimum. OK. An important point, triggers for actions in a way to save money. This is basically the value of information. This is the main mechanism in our value of information analyzers to create the value of information. They need to be connected to actions in a way to save money. And then I think at this point, here came a little too short. It's an SCNM systems. We are relying on excellent SCNM engineering, SCNM systems. And here, data normalization needs to work. So we don't want to find any temperature dependencies or even faults, alarms due to whatever. It's temperature that's a problem. But I think, to quite an extent, the data normalization is working. It may be a challenge when we have very sensitive eye movements than data normalization, is or can be an issue. We need the damage quantification and localization. So this could be, this is a field of research. First, second, the example, the zero bridge. And this may be a way to, and it is a very efficient way, to come to a damage quantification and localization to work with a model with simulations. Yes, we have also been thinking of the sensor lifetime, one point. There's records of sensor technologies like operating wire sensors lasting for 30 years, also final break sensors lasting for 30, 30, I think, or 35 years. But there's a few examples of such long lifetime of sensors. But the point is that the S&M systems, almost all S&M systems, maybe not 20, but it was 10 or 15 years, should be available in the market for the purpose of the subpoor infrastructure integrity management. And I would like also to add here an innovation perspective. Actually, I wanted to discuss this yesterday in the discussion, but for some reasons it was not possible. So when we look at formations of S&M or looking at how S&M systems are developed, then the developers are going after the detection of very small damages with a high confidence. So they are thinking in terms of reliability of their systems. This is often very much important and it's the right way. And especially in the aviation industry, this is a requirement for the S&M systems that they can detect very small damages with a high probability and a high confidence. But this is a little too short-sighted. Especially if we look in the context of value of information, I think the context we are providing here, this is the perfect ground for innovation because we need to come here from the thinking and reliability for S&M systems to come to think in value of information. So an efficient S&M system may provide a tremendous value of information. Like the first statement in the presentation yesterday, there are tremendous benefits, but only together with the decision scenario, with the structure, and with the actions which we create. So in this sense, the value of information analysis are the ground to basics for innovation because it opens up not just to think in reliability, detection reliability, but to think in value of information. And there's clear ways just in S&M engineering to just do the engineering not in the next sense of analysis to go for a very high value of information. OK, so this was a day one. The points of decisions of or about S&M should be taken in a design phase. What, when to measure that, with what technology is important? The follow-up of the measurements by designers triggers for actions in a way to save money and the S&M systems development, which we are, and availability, which we are relying on. And another perspective, I think we had in the panel discussion on how we get value of information analysis and I think this is three points. Also in the order here, one, two, three. The first most important point is the documentation and dissemination of the benefits. So I think this is simply the most important. And this is what we are after with the case studies. And if we can show very high benefit to different stakeholders and stakeholders of the industry, world society, or whether all together, then this can trigger developments, including influence on standardization. Of course, and this is also the action about we should provide material on how to implement guidance standards and how to actually measure. And specifically the codes and standards, and we are seeing this in the QR code standardization. We need an opening here for such value of information analysis or decision analysis. This is most important. And I think a very good example, which we have yesterday, was the oil and gas industry. What happened in the late 90s? The codes and standards opened up and they didn't state an ethology, but they simply said that we need to have fixed inspection levels, but they need to be justified. And this basically triggered the developments of the risk-based inspection plan. OK, I have my hands for now. Let's have a good day too. Thank you.