 Yeah, just occur to me that I almost have a 10 years anniversary anyway. Yes, as I mentioned I'm a staff member at the European Spalation Source. And I can change this slide. I can't. There we go. And the tone of this presentation here is science and data analysis at the European Spalation Source. I've seen somewhere data reduction also has been mentioned and that's okay because I will also mention that briefly. So first, just to put this presentation in context, let me mention two Swedish events. The one is in, or was in 1994 in Stockholm where the Nobel Prize in Physics was awarded to Schul and Brockhausen for demonstrating that neutrons can be used for probing the atomic structure and dynamics of materials. 20 years later in Lund, construction of the European Spalation Source began. And it has a construction budget or had a construction budget of almost 2 billion euros. And this is just to say that society actually thinks that Newton's scattering is quite important, otherwise to make that kind of investment. So for this presentation here, I will give a brief introduction to Newton's and Newton's scattering. Talk a little bit about European Spalation Source and then reflect him from where I'm from, discuss scientific software for instruments and users. We also like to discuss open science and fair data because I believe this will have huge impact on how science will be performed in the future. And then finally, if you want to learn more about this topic, I will give you some links to other resources. So with that, neutrons should probably know something you can find in a nucleus. But it was first discovered in the early 30s by James Chadwick, who actually proved its existence. And I can see now that there's some mess up with my slides, but I hope I will just try to proceed and see if it works. But it was first published in or he first announced that there was that the neutron possibly existed in a letter to nature in 1932. So what is unique about the neutron? It is, first of all, that it is a neutron, hence the name neutron. It is a spin, a one and a half particle. It scatters from nuclei, so steeply penetrating. One way to think about that is that the distance between two nuclei is in the order of electrons, whereas the radius of the nucleus is in the order of two meters. So seen from the perspective of the neutron, there's a lot of empty space in a crystal. Whereas if you are a photon, the space is full of electrons. It's basically just one big electron cloud. So that is steeply penetrating. But it can also scatters from magnetic dipoles, if you have unpaired electrons, due to the spin of the neutron. And then also quite importantly, it actually has a non-zero mass on like the photon. And to the right is a figure from a famous booklet that you can download from the internet if you want. It's a good starting point for learning about neutron scattering, but it shows the different interactions for neutrons. You can scatter from the nuclei. You can have dipole-dipole interaction. There's an unpaired electron, whereas the x-ray is scattered due to electromagnetic interactions, or due to interactions with the electrons, and electrons due to electrostatic interactions. So that's the different types of interactions for x-rays and neutrons and also electrons in this case. The question is then, no, I just want to first to mention, I mentioned that it has a non-zero mass that's important. If you use de Broglie's wavelength equation, this one up here, in combination with the classical expression for the momentum, it's fairly easy to show that if you can measure the velocity, you can also determine the wavelengths. And this is actually very useful in practice and for many techniques, because if you can measure the time of life from when the neutron was a point A and when it was stated a point B, you can calculate the velocity and from that you have the wavelengths and the energy. And that is a very important feature for many techniques in neutron scattering. So that's a way to measure the energy or the energy loss or gain by a neutron. The neutrons we are interested in have wavelengths in the order of angstroms and usually we call them thermal neutrons if the wavelengths is between one and four angstroms and cold if the wavelengths is longer than that. And basically because it corresponds to thermal temperatures and low cold temperatures. So that's that. The question is now how do we get hold of these neutrons? So basically, Chadwick just proved that they existed. Now we want to make use of them. And one way to get neutrons is from fission. So shown here to the left. A neutron hits a target nucleus. It leaves the nucleus into two other smaller nuclei and in the process it also generates three neutrons and each of those neutrons can then hit another target nucleus and then again create three other neutrons and then we have a chain process or a nuclear bomb. But it is possible to control as we know the process so we can have nuclear reaction and it is also possible to extract some of those neutrons from the reaction. And this was proposed already during World War II by Wannan and Schull. I worked on the Manhattan Project and part of that Manhattan Project was the X-10 Graphite Reactor at Oak Ridge National Laboratory and his idea was then to be able to extract the neutrons from that reactor and use it for starting the fraction from a single crystal. And the latter in the middle is actually to the, I don't know exactly who, the director of some kind that is asking for permission to use an opening in the pile as it says. At least this work could be done. And they made a lot of very, very interesting or got a lot of interesting results and Schull actually also got as I mentioned in the beginning the Nobel Prize in 1994 for this work. Just an example is the loud fraction from sodium chloride crystal from 1949. So neutron scattering is really a side product from a good one, I think from the Manhattan Project. An alternative to fission is spallation. So this is what will be done at European spallation source. In that process, protons are accelerated, hence we have this long accelerator up to very high velocities before they hit a target of the neutron reach material that will be sitting here in the center and that is composed of tungsten. And that then knocks out, in very simple words, it knocks out neutrons from the nuclei and the neutrons are cooled down by supermoderator and then guided through these lines up to the instruments. And these for these long instruments, these guidelines are approximately 150 meter long. So it's very big instruments and the sample that you would like to study at ESS would then be the place of sample environments at the end station here or over here or down here. So that's the spallation process and different from the fission. You need to keep it going, you need to add energy in order to keep it going whereas fission is a chain process that can explode. So there are some safety differences as well between those two. But I guess the real reason for using spallation is that we can get more flux out of that process than it can get from reactors. What you see here is the flux from different sources over time and actually Chadwick's original experiment is down here, one neutron per second per square centimeter. You have the X-10 reaction here and then we have IL and Chernobyl over here which I guess is the most powerful or at least close to being the most powerful reaction in the world, research reaction in the world. But the flux is basically saturated, it's impossible to get more flux out from reactors. Therefore people have started to look at the spallation process and you can see that now it is possible to get more flux out from spallation. J-Park in Japan and SNS in the U.S. are both fairly new sources and the European spallation source will then be on this plot, will be somewhere out here. So that's the reason for moving to spallation and use the spallation process. This is actually quite interesting, I find, to put this in a global context. I mentioned SNS in the U.S. and J-Park in Japan that they both have spallation sources and we are now constructing ESS in Europe. Just like spallation sources, there's actually high flux reaction source in each of those major economic areas. It's higher in Chernobyl in Europe. There's also a free electron laser in each of those and its European expel has started recently in Hamburg and we have the high intensity X-ray synchrotron which is ESI in Chernobyl. So each of these areas actually have one each of those tools in their toolboxes from Intel Science. So what else is different between neutron and X-rays? One big difference in particular for life sciences and also hydrogen energy and in batteries is that neutrons can see light elements. X-ray, on the other hand, they see electrons and then hydrogen atoms only have one electron. So it's very hard for X-rays to see the hydrogen atom or the cherubon atom because they still only have one electron. Whereas if you go up in the weight or in atomic number, the number of electrons increases and it becomes a lot easier to see the element for X-rays. For neutron, hydrogen turns out to have the bigger cross sections of all the isotopes because the scattering is isotope dependent for neutrons because it's scattered from again from the nuclei. So the scattering cross section is actually different for hydrogen and deterium. But you can see the light elements with neutrons and that is shown here to arrive with these old-fashioned analog cameras. In the top you see a picture radiograph just like the one you get at a hospital with X-rays just using neutrons here and you can see that the middle casing is basically transparent whereas you can see the organic material, you can see the film inside it. For X-rays it's quite the opposite, you can see the middle casing but you cannot see the film at all. So this is how the two techniques complement each other. There's also this famous movie on YouTube where you can see the expression maker in the work. You can find it on YouTube if you want to see it all but it's basically the expression maker is transparent, it's aluminum whereas you can see the water inside it. So let's stop that. Okay, so neutrons versus X-rays they complement each other. So neutrons see light elements, particularly hydrogen but also lithium which is important for batteries. They see unpaired electrons, so magnetic materials, they're deeply penetrating, isotope dependent and that can be used for something called contract matching. I will not discuss that here I'm just mentioning that you can by adding heavy water to actually a solvent you can actually make a component of the solution invisible which is a quite useful technique in life science. It's also non-destructive, it doesn't destroy the sample that is being studied. There are also some not so good things otherwise it would have been more widely used I think. It's very low flux compared to X-rays, it means bigger samples and longer exposure time. I think for instance for protein crystallography you need a week's exposure time at ESS. So that's quite different from synchrosone. It's expensive and there are very few sources available. So that's the difference between using X-rays and neutrons. European spellation source next through the next floor. We have the bridge and we have we can't see it here but something called the data management and software symbol over here in Copenhagen. I'll come back to that. But ESS this is a rendering of how it's a vision to look like once it's fully constructed which will be pretty soon. Right now it looks like something like in the photography here to the right that's from April and you can see the buildings are popping up and actually the first staff members have moved to one side and now the day they worked there. In 2014 when we started there was a complete empty field so called Greenfield and nothing was there. On this picture you can actually start to see the contours of the accelerator and the target building and experiment holes here. And all of this is going on in Lund of course. ESS is a spellation source but it's quite different from other spellation sources in operation. It's a so-called long pulsed source. That means that the proton being hit is the target 14 times per second and for each time the pulse of nutrients is generated and at ESS that's a long pulse whereas at other sources it's a very narrow pulse compared to ESS. And what it means is that the time resolution mentioned time of flight before is not as good as at other sources unless you make the beam lines very long or the longer distance you make so the time of flight the more accurate it gets compared to the lengths of the pulse or you can also chop up the pulse using choppers that's basically rotating wheels that chop the beam off and in that way you can generate a series of very narrow pulses. So you can basically shape and design the pulse as you want to and that comes with a number of advantages for instrument techniques. And of course the long pulse also means you have a lot of flux and intensity compared to the other sources. The toolbox at ESS is we have 15 instruments planned. There's Odin for imaging, the defraction we have an instrument for engineering defraction, one for power defraction or two for power defraction, one for single crystal, and one for macromolecular photography. For large scale structures we have two instruments planned for sands smaller than the militant scattering and two planned for reflectometry. And for spectroscopy we have one planned for molecular spectroscopy that means that vibrations for quins quasi-elastic militant scattering that's for instance diffusion or rotation of missile groups that you can detect that. And in elastics and scattering that's phonons and spin excitation, spin waves, etc. You can prove with that technique. So there's a wide range of different instruments designed to study very different materials and phenomena. And just to show the scale again this is a rendering of the instrument Odin. That's an imaging instrument. It's located here and hard to see but there's actually a person standing there and that's all like if you're a size of this instrument here. It's big instruments compared to x-rays facilities. A lot of concrete. So what it means is just like for synchrotrons there is a wide range of time and link scales that can be studied with this different scattering. Everything from physics, electrons and tigre seconds down to engineering and material science. And I'll just give a few examples. Imaging, there's computer tomography, standard technique I believe in x-ray. But when you take radiographs from different angles and you can combine it into a 3D volume. And also on the slide here I have an image of a fuel cell but you can see how water is distributed. Again because you can see the light or you can see hydrogen. But going back to the computer tomography that can be used to study how a cannonball is constructed. This is from the battle. I don't remember the year but it's from the battle in England many years ago. And this is a study by under Kastner at the Polchier Institute in Switzerland. And the green color is lead. And you can see some rust because inside it there is a core of iron. That's actually asymmetric so I guess this cannonball with tumble in the air and then probably is a lot more dangerous in that way. So this is just one example in one extreme end of how you can use neutrons. Another example is actually a bit related to this is another microscopic object. This is a railway rail. And for each of the dots here a diffraction measurement experiment has been done. And from that you can calculate the lattice constant for each point. So what you can see is that in this railway rail there's a lot of residual stress in particular the red color part here. I guess the lattice constant is shorter than usual. So that's another example in the macroscopic world or link of macroscopic elements. And single crystal diffraction you can study or you can obtain both the crystal structure of the crystal but you can also obtain the magnetic structure again of this crystal using neutrons. And again mentioned you can see the hydrogens in proteins and that's of course very important because proteins are a crucial part of the function of enzymes. So that's a very important study and there are a lot of expectations to the instrument NMX. And finally spin waves in uterum iron oxide can be studied for instance at an instrument like B-frost at ESS in diffusion. And I should say nothing to do with these examples here except that I have found them and now using them for this presentation here. But it's just to short the wide range of studies that can be done at ESS. So scientific software for instruments and users. Again at ESS we have the data management software center and that is actually not located in London but in Copenhagen. And this is where I have my daily work outside Corona times. So we are located in in this Covis building which is on the campus in Copenhagen and you might have seen the building for if you collaborate with somebody at University of Copenhagen. But our task at the the DMSC is simply to provide all the scientific computing or scientific software that the users need and the instrument teams need in order to to perform the experiments and analyze the data afterwards. The way we look at it is through this data pipeline. You control your experiment and you acquire some data. You can stream, it's events. You can basically detect each time in Newton hits the detector. And we also stream mesa data like temperature, pressure, whatever. We reduce the data meaning we convert the raw data to density basically as a function of the scattering vector and energy gain or loss. We analyze those data and that's technique dependent. And of course we also, let's say of course that's not always the case, the synchrotons are no but we restore the data so we later can retrieve them and we use the next format for that. Also the scope of the data management of software center is through provisioning all the hardware required to run this data pipeline. And we also respond to the user office software. This is the software used to communicate between the users and the synchrotons. So that's the data pipeline. Just to show an example, so down here we have the detectors, we have the raw data that is not animated, I should say, but you can see where you have a high intensity on the detector. You take those raw data and you convert them through basically histogram. So intensity is function of the scattering vector. This is a sansage permit by the way. So we just have the scattering vector here. And then you have a hypothesis in this case the scattering is from a sphere. And we actually know how the scattering pattern from a sphere looked like and how it depends on the radius of the sphere. So we can fit the model to acquire data or the reduced data. And this is being updated as data coming in. And you can see here this is the radius and you can see that it converts towards a value which is slightly higher than 25 angstrom. And also here the background is also fitted in this process. And this is actually a typical workflow for analysis. It will be demonstrated at the two next webinars for diffraction two weeks time for now and sans in one week for now. But basically acquire data, define a model with some three parameters and calculate the scattering pattern for that model and fit the three parameters to the experimental data. It's a typical workflow but it's not the only one. There are also model analysis that can be done. So another thing which can be done, this is just to go a little bit outside the textbook examples is that we can take an existing model for water, a force field, used for MD simulations. And for those of you who have done MD simulations of water, you know that there are many, many different models for PC, SPCE, TTP, etc., etc. And we can do MD simulations for each of those models and we can calculate how the quasi-elastic mission scattering spectrum will look like. So this is from an MD simulation and this is the real data and these are actually from Bertie Halle at Lund University. So this research is actually an example of how data can be reused for other purposes. And we can see, so I guess this is tip 3p, that there's quite a big difference here. And again what we can see is that tip 3p is actually an outlier and does not reproduce the dynamics very well. And it becomes worse as we reduce the temperature actually down to supercooled water. So this is just an example of how a quince can be used. Just for evaluating the existing models. This research is, by the way, funded by the Research Council and it's in collaboration with the ISIS, where Thomas Farmer is, Anders Magmustender, and Heluisa Bordala at the University of Copenhagen, and Jens Svindsson at Thalmas is the PI. And I'm also involved in this project myself. Anyway, we need to support all techniques for data analysis and that's a huge task, because there are really many different techniques and there are really many different scientific areas. And they all would like to have different software. And we cannot do that ourselves. We simply don't have the resources for doing that or expertise. So we collaborate with basically all the resources in the major ones in Europe and also in the U.S. on that. So next week there will be a presentation about Sashview for Sands. There's something we do in collaboration with ISIS in the U.K. and I live in France as well as S&S in the U.S. and NIST. And other facilities that's really a big international collaboration. You're also in two weeks from now here about East defection, Fultraff-Krispire for defection. And that is in collaboration with L&B, Iowa, France. So we do collaborate with basically the entire world about techniques. We also collaborate with universities about developing new missiles for analysis. And I should say part of this work has been funded by the EU project called the S& 2020. Good, another. So one challenge is of course the diversity of the science and the techniques. Another challenge is the community, which is very diverse. Some maybe the most specific software is Excel that I used and for other people like to script an entire experiment using Python or Meta. And we need to support both types of users and the typical way of doing that or at least the way we do that is to provide those graphical user interface and scripting interfaces. We do have in order partly for our own sake, because we operate with a lot of different software and we would like that to look as much as possible. So it's easier to move from one to another software. But we would also like to make it easier to collaborate with other facilities. And in particular, that for instance, scientists at universities also can contribute to data analysis software. So therefore we have a set of standards or guidelines also developed in the project signed 2020. And basically it recommends that you see plus plus if you need something high performance language, but otherwise Python, particularly for scripting, if you want to use make a GUI, use QT or try to avoid it, would be my recommendation. And I can also say that all the software we work with is on GitHub and open source. What I would like to stress though is that it's extremely important to do testing, make it a happy to write tests, and also version control. And it's simply a matter of developing maintainable and sustainable software also for your own sake. For those of you who might want to develop software for research or maybe even contribute to existing projects, I would recommend to participate in a workshop on how to develop research software in a sustainable way. And there are different places where you can do that. But one is the code refinery that actually organizes workshops at universities all over in all the Nordic countries. So that's just a recommendation from my side, because I think software becomes more and more important. And I think most scientists these days sooner or later gets involved in software development at some level. Yes, it is becoming an increasingly important just an example. From looking at cars, 1977 actually has software in it, 50,000 lines of code. But in 2015, she evolved that has 10 million lines of code. And then how you all find a different number that says 100 million lines of code for a modern high in the car. Some of them are the demand for software is increasing. That also is the case at Lytton facilities. Mantid, this is one of the two software we use for data reduction at ESS. There's two and a half million lines of code and the cost so far has been more than 25 million euros. So it's not cheap either. Also, another thing to pay attention to is that the more lines of code you have, of course, the more functionality hopefully the more science you can do. But it also comes to small complexity and therefore an increase in risk of failures. So be concerned about whether the software you're using and the data you publish actually are reproducible. There's another thing that's coming up these days being discussed. With the ESS, we will have instrument data scientists that will assist users with data processing. And during a construction phase, they will also facilitate that we at DMSC and the RR that you can see, we actively work with software to the instrument teams. So open science and fair data still have a bit of time. This is being strongly pushed by the European Commission. Basically, all large scale facilities in Europe, I think at least the international ones my ESS have committed to open science and the European open science cloud. And one important component of that is the so-called fair data and associated analysis services. Fair data means that data are findable, accessible, interoperable and reusable. And what does that mean? It means that, for instance at ESS, you go and do an experiment, then after three years, other people will be able to use those data. That again means two things. You need to document to provide a sufficiently high level of metadata. So other users actually can, I mean, not only access the data but actually also can understand them. At least if you want to, then to understand. But you also need to make sure that your data are sightable, so that they have digital object identify. So that's very important. And actually, before I gave an example of how open data can be used, I should be honest and say in this case, we just emailed Helen and got the data, but it is an example of how data can be reused for other purposes. We currently, ESS is a member of a project called Panosq together with the other so-called H3 on the European, what is the strategy for research infrastructures, so the international facilities in Europe, so Excel, ESIF, ELI, European laser infrastructure, IVL, and Siri Eric, which is an umbrella organization for smaller research infrastructures. We are implementing fair data policies, and we are also implementing techniques that, if not exactly sure, and these facilities that, for instance, analysis can be reproduced. And one of the techniques here is to use Jupiter notebooks that I think many of you probably know, because Jupiter notebooks can be stored as an ordinary data set. So as that was just said, that is not only the international facilities that is implementing open science and fair data. We have a sister project called Expand, where national facilities are members and that includes Max4. So Max4, we collaborate closely with all these facilities. So this is really a European wide process of implementing open science and fair data. Data management, since this is a link seminar, I think it would be important to mention that we develop a meta-data catalog jointly with Max4 and also called the Institute in Switzerland. It's called SciCAD, and so that would be a place where you can search on meta-data to find specific data sets that you are interested in. Good. Just a bit about if you're interested in learning more about Newton's scattering or what can be done. I mentioned the Newton's scattering primer by Roger Pin. You can download it from the internet. It's a very good introduction, but I'd also like to mention panlearning.org slash enurtron.org. So panlearning stands for photon and neutron learning. This is an e-learning platform. Originally, it was the enurtron.org because it focused entirely on Newton's scattering. But as part of this panors project, we are now expanding it to also include teaching material for photon scattering or for x-ray sources. For instance, you can find an introductory course to Newton's scattering in there. It has quizzes, textbook materials, slides, videos, and you can also, and as part of teaching materials, you're supposed to do simulations of virtual instruments. So that is actually very good. And I think we are interested in the project in collaborators if somebody wants to help this putting context into this platform. That's both for the photon side and for the neutron side. I'd also like to mention that if you need some specific software and you don't know what to use, there's something called the Pan Data Software Catalog, also now maintained by the Panors Project. So this is a place where you can search for software for doing sands, for instance. And it relies on people registering software themselves. So it's not fully comprehensive, but I think it's actually fairly close to be it. Finally, the next upcoming webinar is one by Wojciech, my colleague, and my other colleague, Andrew Sassanoff. So the next one will be on small-angle scattering data analysis with Sassview. And I guess that also will include sacks, meaning x-rays, and also an easy-to-effection that will be two weeks from now by Andrew Sassanoff. And that is a piece of software which has been developed entirely so far at DMSC and links to other software like Fulproff. I think by that, just in conclusion, or as you mentioned, there are also several summer schools on this topic, including Sweden, of course. So yeah, neutron complements x-rays. They don't replace them, but they complement x-rays. They're sensitive to light elements, magnetism, and deep penetrating. I also like to, modern scientists need software, and it's really good if they contribute, but be careful with sustainability and reliability. Fair data and open science is becoming a reality. And finally, you're welcome to contact us if you would like to contribute to analysis software. And then, of course, I should also mention, importantly, that the European's medicine source is becoming a reality as well. I think that was all I would like to say. So yeah, thank you for listening. So if there are any questions. Thank you very much, Thomas. It was a really great talk. We collected some questions. Yeah. I'm going to read them to you. So the first question is, is it possible to monitor organic reaction or bond breaking using neutrons? Indirectly, yes. I mean, catalysis has started a lot with tumor analysis, molecular spectroscopy. So basically, it's the change of vibrations in what you are looking at. So Tosca, that ISIS is an example of that. So yes, indirectly, I would say. I mean, it is used for that. Okay. We continue with the next question then. Can we do time-resolved experiments to study photo-trigger reactions or to detect the transition states in the reactions? Pass. I don't know. But you can study time-resolved reactions. Yeah, I don't know. I mean, for instance, next week, I'm sure that VyTek will talk about time-resolved sands and sacks. But I mean, there is a limit to how much you can do due to the limited flux, I would think. But I don't have a wise answer to that. I could find out. So as I said, the audience has different backgrounds. So the question can vary a lot. So some of them can be specific, others a little bit less. So we have another question. So you talked about the importance in your presentation. What is the meaning of sample volume? And concentration for your neutron experiments. In terms of how well it scatters? Yeah, I believe so. Yeah. Well, I would think that the higher density of nuclear power you have, the more scattering you have and the bigger volume the more scattering you have. You need fairly, I mean, compared to exploration, you need a fairly big volume, so sort of a beam size. Typically, we talk about centimeters. As a synchrotransmitter, we talk about million meters, right? Or even below million meters. But yes, it has an impact for how much scattering you get out. Of course. So another question is, can we study the interfaces of hybrid devices using neutrons? Yes, you can. I mean, I guess it's layered structures. So we have a reflectometry for that. So that can be studied, yes. I know solar cells is one example of being studied with neutrons. Okay. The more we talk, the more questions we get. So we're going to use you. So typically, how long does it take to acquire data at the neutron beam line? Anything from, it depends on the cynics. So for instance, for powder and science, we have higher flux. And it also depends on the facility how much flux do you have. I mentioned in the next talk about leaks. I mean, that's not the normal case. But I would say anything from hours to a leak. So it's really, it really varies between the different facilities. Yeah. I mean, it depends on how much flux you have, of course. So ESS would be more powerful. It means that you can do, you can do better time-resolve experiments, which was mentioned, right? But you also need shorter exposure time, surely. Yeah. In fact, there is another question here. We learned that X-rays can damage a sample. So is it the same case with neutrons? No, it's not. Neutrons does not destroy the sample. That's one of the advantages, yes, of neutrons. Great. So... I mean, there are some elements that becomes, I don't even think that destroys, but there can be elements in that becomes radioactives, like can be in stuff like that. But yeah. But usually, no, it does not get destroyed. Okay. Another question is, is the data reduction performed on-site by the user? So I think... I'm grateful for that question. Yes. Actually, I did not mention that at all. Both data reduction and data analysis is performed on-site, meaning that the computers are on-site. You don't have to take the data with you. You can do it, but you can log in from home and do data reduction and data analysis. You can also do it on-site, of course, during experiment time or beam time if you want to. And as I mentioned, we also have an instrumentator scientist who can assist, particularly through this very complicated analysis. So this would be a DSS, or this is valid for all facilities? Oh, this is for ESS, but it is actually also the plan in panels for all those facilities that are members. And it is up running for LISIS in UK, that so-called ITAS system. And also it's an ESS to have some remote analysis. So you can log in remotely and do your analysis on-site. Well, you don't have to be on-site, but you don't have to take the data out in order to analyze them. And that's simply because the data sizes are becoming bigger and at some point it becomes impossible to move the data out, or at least not feasible. So that's for notion sources. I mean, the problem is even worse or much, much bigger for synchrotrons because it produces a lot more data. Okay. We have another question. This is a little bit more specific. Is it possible to quantify the hydrogen segregated at the grain boundaries in metallic alloys? If yes, was the lowest detectable limit to one kind of ship? Yes, that's too specific for me.