 A warm open group in Dublin, welcome please for Johan Crevers. Thank you, thank you. Good morning. So we're going to talk about really what is ODE all about. Why do we need something like an open sub-serviced aid universe in the oil and gas industry? And also what's all about the role of open group and where we are and how we're moving forward. So it's really if you stayed from a background and where we are and how we're moving forward in this space. Let's first give a little bit of history. If you look at a typical oil and gas company, it's much in the subsurface space. Subsurface is really the space where you try to find your oil and gas futures. Subsurface is really where you go out there shooting your seismic, collecting that data, go into the development phase and try to get wells out there to start producing it. So that whole part today is very much in a siloed environment over there. So you've got data linked to applications, the first one. So you've got many, the data is very intrinsically part of an application out there. Therefore data is stored in many silos. So there's not really a lineage out there. What does that mean? A lineage means if you're in a well space, I want to understand why we decide to drill a well at a certain place. They go all the way back to the exploration phase. It's almost impossible nowadays because the data sits all in these silos. It's very hard to figure out, okay, I've now decided to drill a well out there, but what is all the information that led to that decision? That's very hard to find out nowadays. Sometimes you wonder how you still do find oil and gas because of all these limitations out there. So data is stored in silos, very important. And then metadata is stored with the data. Metadata, of course, is data about data, but often we have millions of files out there in the subsurface space with many different file formats, but very poor with the metadata. It's very hard to find out what data we actually have. If you look at your seismic software, if you look at your interpretation software, what actually is the data we have out there? So we have lack of content information. If you sit, for instance, in the case of Shell in Australia, and you want to know what we're doing in the US, it's today's impossible to find out. Limit to no search capabilities. It's very hard to say, okay, because there's no metadata, it's also very hard to say, okay, but what's the search, what data do I have, what kind of fields do I have, what kind of common fields do I have around the world? Again, very hard to figure that one out. And the last one, which is very important, data is not ready for a data-driven world, because we're really moving from, of course, a physics world to physics and data-driven world out there. And data, because if you look at our data today, only maybe 10% of our data is actually being used in the subsurface space. The rest is not being used, the data is there, but it's not being used because the people are not around to make it happen. But with machine learning and AI, it becomes possible that all the data is being used all of a sudden. And that's why this whole part of data application is very important. So this is the world of Shell today. So I'll speak here now for Shell, if maybe extra mobile and BP or to tell whatever they're called, do a much better job out there. But based on experience with the OSU experience, we all have that same problem out there. Lots and lots of data. If you look at a company like Shell, in this space, you find by far most of our data. In our case, it's about 200, over 200 to 300 petabytes of data sits in this particular area. And many different data types, and not all structured data. If you go to all the parts of a company, often very structured data-driven, all over the place. You've got structured data, unstructured data, binary data, many different formats out there. So it's quite a complex environment out there, the most complex environment we have. But the current reality is also that we're not in a very good shape out there. Therefore, we cannot just continue because we are moving to a data world out there. And as long as you have the data sitting out there in your silos, it's very hard to bring the data together and say, oh, now we've started applying machine learning in a way that we don't know what the focus is. That's the first point, because we move from physics to data, and data world. The timelines we have to reduce. The times today are far too long to go from shooting your seismic and really drilling a well out there. Data quality is not always very good, needs to be improved. And of course, the access to metadata. Metadata is highly underrated, but so important because it gives you information about data. And metadata is very hard to find out what information you actually have. What makes it a challenge out here, that you have so many different file tabs out there that the metadata, you need the metadata to extract from every file tab out there. So it's quite a complex environment. So we cannot control like this. So the biggest change we have to approve is the first one. It's really separate data from application. It starts from a data point of view. Forget about the applications for now, just get your data store in order. So put all the data into a separate data platform and separate the data totally from your applications out there. And such that any application can access any type of data. But start from a data point of view. And look at all data. That goes from shooting your seismic to a pre-stack, the post-stack, your drilling, your well logs, your borehole data management, any type of data out there will go into that single data platform out there. And the scope we're talking about first, and that will change over time, we're talking first about exploration. Exploration is really the first phase. You're going out shooting your seismic and do something with that seismic data. Then development, you've decided that you're going to do something. You're going to start developing now your field out there. It's development phase. And well, especially with the phase, we start drilling your wells to start producing out there. So that's the whole phase we're talking about. Together we call it subsurface and a well space. So data descent and support for metamaster data. Making sure that when you're loading your data into a data platform, that you do extract all the metadata out there. That is gold for when you start looking for information about the data out there. And of course real-time. Be more of a real-time company out there. Also in your subsurface space with your reservoirs, we're more of using sensors to real-time collect data about that field. So it's not just a batch environment. More and more comes also a real-time environment out there. But the first point is to put data at the center of what we're trying to do out there. This is the most important point we need to do. And for the rest, applications will fall. I'll come back to that. So data descent and any type of data. If you look at this, the overall high-level picture of what we call then the open subsurface data universe is both a data platform and an application platform. And I'll play the calling in a moment. But let's first look at what we call, maybe sort of a green calling. It's a send you find that the data platform services, that's where you store all your data. So all the data we're talking about, all structured data, unstructured data, your documents, whatever you have, is loaded via the data ingesting services into your data platform. Your data ingesting services are very important. But that's where you extract metadata. So you have your data ingesting services. That's where you decide for each data type, your file type, what's the metadata I can extract. And the method is, again, is the information about the data itself. There's a data ingestion layer, you extract your metadata, you also look at the data quality, and you bring all the data into your data platform service. So all the data you can think about, whether it's your raw seismic to your drilling logs at the other end, will go into that single data platform. Of course, this all sits in a public cloud-based environment. I'll come back to that later. It's a public cloud environment. And of course, you also have your information security out there, which has three layers. First, you use authentication. It's all about who has access to what. User authorization. If the user logs in, what type of access do the data, does that person have? Does it run by roles and by metadata again? And thirdly, of course, the protection of the hosting platform itself. And then, of course, you've got the whole support environment. You're going to support the management. You operate your order trails, you're monitoring your CSID development environment, et cetera. So that's how much the green part. And then they also find the green part in country, because we also operate environments where we're not allowed to move the data out of country. But if there's a public cloud environment, public cloud is not available in every country in the world. And we have countries out there where we're not allowed to move data outside country. So we have to accommodate also what we call an in-con dissolution. It means I can put a hardware in country to support that same capability. That's our in-con dissolution all about. It's that in-con dissolution of your public cloud environment. You see here, the green part is the part, so now we're going to talk about the OSU environment, as part of the open group. The green part where we as operators do not compete. I do not compete with BP, with Total, with ExoMobil in the green part. We do not compete our structure data. We will compete about the data itself, but how we structure data, who cares? Nobody would care about that. So the green part is clear part. We can work well together. And as part of OSU, we are working well together. I'll come back to that where we are on that. The gray part is where we do compete. Those are the applications. Because everybody will have their own set of applications, either developed in-house or budget on the extoller market out there. And the red part is a clear separation, the API layer, the application program interface. That's a very important thing out there, because that determines the separation, but also tells you what the applications need to do, how the applications will get access to your data platform layer. But what we try to achieve with OSU is a platform capability. They have lots of parties who develop in the gray part, whether you are a startup, whether you are an operator, whether you are a large company, a small company who cares. But every company will develop applications in the gray part, so we create competition out there. I'm going to explain in a moment. Therefore, the API layer is very important, because it gives a well-defined endpoint which you cannot just change for people who develop applications. And whether we got OSU installed at Exxon, or in Shell, or in BP, or in Woodside, or in Petrobras, we don't care, because that application will run in each of those environments and the API is always well-defined. Of course, different data out there, but the API is well-defined. So all this is public cloud-based. There is public cloud, public cloud, and public cloud, so no private cloud in this space, apart from the income dissolution. So all data is loaded by the data ingestion services, where we recognize the data types, the exact metadata from the data types. We have the set of RESTful APIs out there, all applications. We support both legacy applications out there and new applications, because our new applications we're all talking about will all be HTML5, Microsoft-based, but of course, we have a large number of legacy applications out there. We still, often, are still Windows desktop or Linux desktop-based, and we support these as well, because we have to move the migration forward. We talk about security, we talk about in-country, and of course, we talked about the legacy environment. So if you're talking about OSU, it's both the data platform services and the application platform services, and it comes for us, really, the single center of truth over there, because all the data, all our subservice and wealth data will be sitting, will go move into the single data platform with all the applications sitting on top of that. So that's the overall picture we're creating. So it covers all our guest companies, although there's very little unique to this picture. I mean, if you say, I want to use OSU in production, what we have to change is very much the data ingestion service and make sure that all the data ingestion services do support the data type being used in the production space. So there's very little unique here to say, well, this can only work in subservice and wealth. There's nothing there. The only thing which makes it unique is the support for data types out there. Subservice portfolio includes X-rays and development of wealth, and of course, we expect this to grow, most likely also into the production space. We now focus on this because it's the most complex part. We need to get it right. All data types, structured and unstructured. Unstructured in this space is very important. Unstructured could be your PowerPoints, could be your Word documents, because often in X-rays and they use, they put images in PowerPoints and in Word documents about areas of the world, with sort of coordinates. But if you use a typical search environment in Microsoft, in SharePoint, it's unable to find these images inside Word documents. Just recognize them. That means we lose a lot of data in this space because we're unable to find that. So we're also loading that type of data into this OSU environment where we can extract the images of those PowerPoints and Word documents because there's a load of information in the environment which today we're unable to absorb. Support for data-driven applications. Very important. Machine learning, machine learning, machine learning, machine learning again. Why is that so important? Because the first time now we can exploit every byte of data we collected. And if you need to realize, whenever we do a survey out there, we're collecting petabytes of data per survey. And we do many surveys every year. So there's too much data for a normal human being to comprehend. And therefore using machine learning to go through all of that data and start steering the interpreter to say, okay, focus here, focus there, and focus there. So start steering, primarily managing, driving that interpreter. Therefore, that data platform for machine learning is also very important. We talked about metadata. Never underestimate the importance of metadata. We talked about the single-cell definitions and of course the API layer out there. So OSU coverage. If you now look at how that would look like, take this shell environment. So although you could say, well, in principle, you could put OSU in one location. Let's put in Dublin because quite a few, the Amazon data sits over here in Dublin for Europe. So you could say, okay, let's put OSU for shell over here in Dublin. Technically that could work. The problem, of course, we have many so-called 3D applications. A three-dimensional application. That means that the distance between your data and your user can only be a certain distance because otherwise it won't perform. Therefore, you see the first bullet point over there, the OSU location driver makes this between use and application, often 3D data. And so normally it's about more than 40 or 50 milliseconds. That means you're talking about between here and maybe 44 milliseconds is well into Europe, so into the continental Europe. Therefore, what you see over here, in the case of shell, we will be implementing them, we will be implementing them around the world. That means every system will have the full set of metadata, so metadata is replicated around the world. Wherever you sit in this environment, where you sit over here in Europe, in Norway, or sitting in Australia, you know exactly what data is available around the world in this environment. The real data will stay within the region that's been created, like all the European data will sit in Dublin, all the US data will sit in Virginia, what AWS data is for the US, et cetera, et cetera, et cetera. And where we have teams working together, we will temporarily replicate the data as well. But you should, the reason for this disabuse is purely driven by the bullet point one. It's the max distance you can have because we've got high-intensity 3D applications with interactive reuse. They're not reporting applications, they're people, 8 hours a day are working with that application. So if that is not performant, you get some very frustrated users out there who will shoot you or kill you, whatever you're going to do with you. Data is replicated globally, it's very important. So if somebody in Perth and Australia wants to know, did we ever have this type of field before? Did we have drill in this type of environment before? You can ask that question using search out there, we're using graph out there, and you get the information from anywhere in the world, but replicated around the world is the metadata. RESTful APIs, every system has the same environment out there. And extra data stays in the region that's being created, like in the US or in Brazil or in Australia, wherever it is. And of course, it also includes the full development environment, CSD and development environment for DevOps, so if you want to develop your own application out there, it's a complete system, which includes your data platform, your application platform, but also your full development environment. As part of OSDU, we will go beyond the normal standards. We're not going to say, well, here this OSDU from a standard point of view, we're also working with both Microsoft and Amazon as soon as we Google for them to deliver an OSDU implementation on their platform. And the first release will come later this year and we'll come back to it in a moment. So it's important to realize we go beyond just delivering some standards out there. But I don't believe just leaving standards out there, that will be good enough. We need to make sure that people have something they can implement, they can work with, and they can actually make use of. And I'll come back also in a moment how we do that. So this will maybe set a foreshadow if Exxomo would do something like this, it would most likely be very similar because it depends on where you got your reserves around the world. If you just operate like Atnog in Abu Dhabi, of course you would only have one system in the Middle East. Purely where are your reserves around the world? That really drives this. And everybody will have their own implementation, so we're not sharing them. So OSDU targets very much. And it works as workflows. The workflows, so far the workflows because the silos were very narrow, they were only workflow within exploration or in development or in the wealth space. Now because I have the data available across the whole spectrum, my workflow is also covering the whole spectrum out there. So my workflow can start in exploration and go all the way to wealths. Again, it's very important, because we talked about linux before, if you're sitting in wealths, how did I come to that? How did I make a decision that we drilled it particularly well over there? Now I can go all the way back to express these were decisions made on the way to come to that decision in the end. And also when in your production phase find out that the field doesn't produce as expected, you can go back again and say what is different in the field than expected was when we did the whole phase before that. Kubernetes support, familiar with that, is my file driven. Means I can run on every type of endpoint. It's a fully browser based environment. Also important collaboration because teams often work around the world. The teams work on the data often do not sit in one location, but often work from multiple locations onto that same set of data. Therefore collaboration and virtual reality is quite important. Physics and data, we talked about it. AI, very, very important and the reason I gave you. There are a lot of applications out there. So if you use Schlumberger Petrel, which is very popular in this market, it's a Windows based environment, Windows desktop based environment. Also that is supported in our OSDU environment. So people can still use their so called legacy applications in what we call a lift and shift mode and lift them across the OSDU space. It won't have all the full set of capabilities as a native application will have but it will run in that environment and it will run in the old application and probably in the new applications out there. So what's the biggest impact of doing this? Yeah? So we have all subsurface which is exploration, development and wealth data in a single data platform accessible via single API. Yeah? Yeah? So silos have gone. So I can go all the way back from left to right, from right to left whatever I want to do. Data is separate from applications. The application no longer owns the data. What's really today? If you go to the patrol days owned by the patrol application. That also means I'm storing that data in different formats than today to make sure it's more generally accessible out there. Global oversight on what data is available. Anywhere in the world I can find out what data, what information available about wealth, about activities wherever operate, wherever sit in the world. Yeah? Because my metadata, my metadata, my metadata is available around the world. And the last one we're creating a world out there where many companies start using the same APIs out there so software companies can start developing applications out there which can be used by a wider group of users out there. Because if you look at the subsurface today you have a limited set of companies out there. You've got people like Sloan Museum, you've got Halliburton, you've got Baker Youth you've got a small set of companies active in this space. And because it's a very, as I mentioned before it's a very solid different space out there. What we do over here, we separate the data from the application so the data is being handled by OSDU and the application is now handled by the software developers out there. And of course we've got a set of APIs being used by all the operators so if a startup wants to say I want to develop software for this space it's now can be used in many different companies out there. Like in the past. So the last point is important. The thinking behind all of this to create a market or create a platform out there where people develop software for this environment so more use of it, the more use of it more people start developing software for that environment. That's something to get into. Always compare with Salesforce Salesforce.com you're all familiar with has about 3,000 partners out there developing software for that platform. We won't just get 1,000 out there yet but this thinking is important. The thinking behind it is what it is. You will have other people that develop for your platform so your platform gets more popular etc etc etc. That's the thinking behind the last bullet point out there. Make it accessible. Okay, so if you go a little bit let's do a little bit history over here. We started with OSDU early last here. So Shell started with the Subsurface Data Universe and so Shell started with the Subsurface Data Universe early last year and said okay for this to be successful it's far better if we can make this industry solution. Because we far better off if many people develop applications for this space so we get more competition out there, more innovation out there. Therefore we said okay together with Phil Biong he's my partner in crime sitting out there so he you can blame the two of us to start all of this but we said okay let's talk to some other majors out there. We did that in March last year. Then we started talking to Open Group in Q to last year. We had a meeting with well we called Steve and said Steve this is what we're thinking about. What do you think and Open Group do? Something like that. Because we need a framework we need a framework to work together. We can't just work together as companies without a framework and the Open Group was that framework out there and with our kickoff meeting in Houston in the Galleria in September last year. When we started in September last year with about nine or ten operators. Exxon, BP, Total, Devon. It was about mainly US based but also European ones out there. It was our kickoff meeting for the OSDU. That's where we are now. That's where we started formally but if you now see where we are this is where our operator members are. So these are people who produce oil and gas out there. It's not one percent up to date but it's relatively up to date. You see all the IOC's out there. We now move into the NOCs and national oil companies. You've got Reliance on there from India. We work with ArtNorq in Abu Dhabi and to get also the NOCs and the smaller non-conventions in the US also in this space. And of course we have several supply companies out there. We have about 45 companies that are going to OSDU since I would say late last year because we only opened up for the non-operators in Q4 of last year. Also notice over here the Amazon, Microsoft and Google. Why is they important? Because they are important for us to deliver the solution into the field. So that's how we went to, we already went in August last year to these three partners to say help us creating a solution in the market which is an operator like Exxon and they come to you and say please implement OSDU for my environment. So these are people either service companies who deliver service into the oil and gas market software companies that are total about about 45 companies today. So we created September last year therefore the OSDU forum under the open group and we created at that time three subcommittees. The enterprise architecture, the data definition and if made security. September last year where the number of face-to-face meetings where the enterprise started defining what it's all about. So we looked at what Shell has been doing with SDU we contributed to the OSDU environment to give them a kickstart and really started finding with these subcommittees the architecture, data definition. Very important was the first time in this history that we aligned definitions. If you go to this industry today and ask them all for what's the definition of a well, you will get maybe 50 different answers if not more. So we're now also aligning that kind of thing. That's the data definition work team do and if made security. We also reached the editor that, the business model to say how do we make the sustainable model over time. This is important also not just to think about what you do the first here it's a final part of how do you sustain is also keep it agile and make sure we move fast into the changing world out there. We, of course, we meet weekly in calls we've got slack, very active community out there and we've got regular face-to-face meetings. Last one we had in January when alone for the enterprise art committee we had about 60 people in the room for three days to define the architecture and the next one is late June again. So it's a pretty active environment out there and we need to deliver something so this was our plans for this year. This is a demo release of OSDU release one by mid-June late June. What does a demo release mean? It's a full copy of an OSDU environment as contributed by Shell a full copy of all the cables we talked about and but it is not yet all the members, all the OSDU four members can get access to the environment it's a single environment I call the read-only environment we don't own data in that environment but we preloaded with data it's quite unique we found 200 wells in the Netherlands of all places which we loaded in this environment so we loaded 2,000 wells in this environment and some demo applications So this being developed by both Microsoft and Amazon so both will offer an environment people can log on to can play with it, can try things out after sometime in June very soon after sometime in Q3 we come with the full release one out there that release one any operator can say like Devin or Exxon or BP or Tutile or Econoc can say can go to either Amazon or Microsoft shop and say please install in my environment under my subscription a copy of OSDU and of course that will be done then and then to get our own environment we first start supporting with wells out there the wells is the first support out there and later this year we also want to add a shadow over here also seismic to this for late this year but the first release coming out is sometime in June and the release after the full blown release for every subscription have their own copy will be later in Q3 So what does that mean this will be a game scene for this industry because this industry has been in the legacy for the last 5000 years and really needs to come out of that and it has been driven by the move to data it's time is everything in this world we know that time is everything with this when we started when Philip and I started this discussion in March last year all these places in the same space if we speak to BP some are ahead some are behind but all in the same space how do I move from a physics world to a physics data world and make data the centerpiece of all that we do so it's a game changer it's desperately needed also we need this environment we need this environment to be successful in the future industry because people say we still need this industry we still need all and special guests for many many years to come to really move to another world out there it's a platform approach what's important there is I have a set of APIs which I make available to any company out there Startup company, Academa we already have like the University of Oslo member of OSDU software for the all-in-guest market as well so the platform is very important to open up that market now that's the thing behind the platform platform platform platform data is at the center to support all AI, ML based applications out there and it's impossible for this industry whether screw me say or sell to solve the data problem on your own it's just not possible people have tried it failed and failed what we do with OSDU we make the data story fully owned by OSDU and we need the supplies to focus on the applications not at the data but they cannot solve the data problem we have to solve the data problem jointly it's very important when we're separating data from applications for software company to focus at the applications, the workflows at the market service out there we already just started we started as a matter of fact I remember last year in Houston I don't know nine months ago whatever it is but we still have a lot of work we come out with first software this year but still a lot of work in front of us drive, drive, drive so join the club because it's quite exciting we're working well together with both the operators and the companies we have a very focused team going on to make sure we will deliver in due and we will deliver in due and deliver for those of us not so familiar with the oil and gas industry without that background can you say a little about the amounts of money and people and data involved in the subsurface world as I said in Shell alone I think 80% of our data sits in this space if you look at the whole of Shell we have what we call an integrated oil and gas company so we've got the subservice all the way to retail because now also the new energy business the home delivery of electricity and what have you so in our case it's between 200 to 300 petabyte of data so it's about 80% of our data sits in that space also if you look at software development most of the software development work goes into that space there's a lot of in-house development still happening in this space because we believe like other companies believe that they have skills in this space unique to that market so also a lot of our software development cost is going to this space and I would say I can't speak for other one but it's very similar for the other ICs out there right, big stuff let's see can you say a little about the business model that supports the oil and gas digital transformation being implemented by LSDU the business model is all about a number of things the first one of course is free of the data because this world like every other world is part of the digital world digital world is driven by data but they have to make sure that the data is available to you and what OSDU does is freeing up the data to make it available to an machine learning environment out there which today if you have all in small silos it's very hard to do it's almost impossible because it was ML about being data together and extracting knowledge of that data coming together so it supports the move from this space from a physics world to the digital world out there it's basically about getting more value for the money that you're investing in and as I mentioned today most of the data we collect we do not read we do not look at because it's too much data out there with this we can look at all the data and then start focusing as you an interpreter and say okay look at all this data you focus here here and here you start by dropping the SME the expert where the focus at okay thank you next question are you making any provisions to ensure the right quality of the data in developing this new standard I always say I always say I don't bother about quality just load the data first but I rectify it but what's important that you don't spend your time trying to solve your data quality issue because you will never solve your data quality issue but try to do in the background in the dark out there it won't happen we tried it for 500 years it doesn't work so we load the data as is of course whilst loading it we have somebody about the quality of data we'll highlight poor quality but the data will still go into the environment even if it's poor quality data and when you start using of course you know what type of data you have and if the results are poor or not what it should be you can also point to who is accounting for the data because you only make data quality better if you can point to people who are accounting for the data and if you try to do it all first improve all the data out there you will never ever get there has the forum considered the use of the open groups open data element framework standard ODEF is that a familiar term to you the foundation starting point for the metadata framework if not there are some people in the room who would like to talk to you let's talk about it today so okay let's see it looks like the OSDU data platform would have appeal beyond oil and gas are there any thoughts at this stage about other industries that might fit or find this useful or obviously an evolution of it useful yeah well we had we already spoke to people in the mining industry for instance and I said the whole model is not unique to this industry what makes unique is the recognition of the data types and data system layer if you extend that say okay I can also recognize production data, mining data all the data types I could also use in the industries our focus first to get this problem solved we will plan is to extend also the production in the same environment so then we go all the way from exploration to producing it using all the industry is not inhibited this purely is the recognition of data type first the whole development the model is nothing in there this can only be used in this industry I smiled when I saw this one this one come in what's been the reaction of the incumbent vendors in the marketplace positive or otherwise mixed of course because if you look at the market today there's one company offers a similar solution out there which is slum is able to Delfi platform but Haliburton is I don't know other people really offering these type solution the only one is what I meant to slum is same but you really want to make sure you have more competition in this market so you really don't want to depend on one supplier out there I guess the bigger companies like the Haliburton will be less enthusiastic at the same time these big companies are not able to solve the problem for us the data problem and in a way we're helping all these companies but say okay the data we make part of the OSDU environment but the application on top of that you as a company can develop at your own less at your own space and what have you okay thank you and last question OSDU enables sharing data access but typically who are the source data providers and what is their motivation for sharing the data most of the data we're talking about is data we either procured or created ourselves because you really talk about the beginning of the seismic environment you either procure data out there or you you created data yourself out there of course we also use external data sources and that has so far not been issued at all but the majority of the data we're talking about is either we procured ourselves or we create ourselves part of seismic we've just had a flood of about six or seven questions all arrive at the same time so we don't have a chance to take them but I'll pick one at random can you elaborate on how you determine what APIs to support the data models in OSDU? we've defined the API search because the APIs cannot change over time anymore because you principally have a contact with those otherwise an application developer will have a problem that the application works today but not tomorrow anymore so the APIs have been defined such that we can keep them pretty static and it really changes in the problem is how you call the APIs so the APIs are fixed but we cannot afford to change APIs in a year from now or half a year from now because then your application developer people have a problem and will kill you I would expect I'm going to squeeze one more in because I can has the front end source see, has the front end source will well head data standard POSC PRSC CISA by an logistics being of any help to your efforts if we look at the analysis from a data exchange point of view and this is also a member of our OSDU and the open group but also part of the OSDU so we are looking very much at this as a data exchange format not as a data storage format but as a data exchange format okay, all right Johan we will leave it there and from everyone at the opening thank you for your passion and commitment and we look forward to seeing the results