 So with that, I'd like to introduce Paul Zeppenfeldt from Shell, who will be giving our keynote presentation. Paul? This works. Good. Good morning, everybody. Good to see so many of you in a real event where indeed many of us see each other for the first time in 3D or 4D. My name is Paul Zeppenfeldt. I manage the subsurface and data and digital parts for Shell upstream. Thank you. It's always good to be in London. I lived here for 3 years myself. Perhaps if you live in London, you don't like London because of all the transport, but I love it. Iconic City, very historic events happening a few hundred metres from here very soon. I promised my family back and all that would bring home some coronation kits, so I'm looking forward to some pillows or tea cups that will look very old in half a year's time. But London is not only the place where in Westminster Bridge you need to find your way through the French tourists as we had to do on our way here, but also where everybody comes together for a purpose. And we're coming together here for a very clear purpose, and that is always the you as a multi-year effort. And it's good before we start today, and I think Steve already referred to how special it is actually the achievement so far. You could really compare the way that the industry has come together for OSDU. You can, in my mind, compare it to things like the OGCI, the Climate Initiative, or even how the industry came together after Deepwater Horizon. So it is really amazing that we're doing this and we're jointly trying to really improve our business. And what's also very clear, and I'll come to that later, OSDU implementation is very much a team sport. So whereas I'm going to try and sketch some of the dilemmas and the kind of progress that we've made, we're very clear that no single company has all the answers. So we're very much also here to learn from all of you and jointly we can make more progress. So really going back to the basics of what we want out of OSDU. I should say that for us OSDU is our plan A, it's our plan B, and our plan C we're really all in. And why is that? So of course OSDU will help us to find the data faster and cheaper. Lots of information, particularly the intelligence out of the information is lost due to handovers or people moving. And it's really a valuable asset. Of course what's important for the data is the metadata to explain the lineage where it comes from, who used this data before, who reviewed it, and it's great that many companies are stepping up, for example colleagues in Exxon are doing great stuff in the OSDU forum on the topic of technical assurance of data. And data finally, we often call it an asset, but you can also call data a product. It is a product that you need to use for the right purpose. And in the end of the day, if you use data for the right purpose, you make better decisions. And that in the end is what it's all about. If I have to summarize what we're trying to do with OSDU in my company, I would always say the ability to make better and faster decisions. Now if we move to OSDU, of course OSDU is part of a whole ecosystem. And you could argue that's the kind of the superpowers that we want to equip our subsurface colleagues with. And if you look from the left to the right on this plot, you move kind of a little bit from left to right to kind of from an area where you want to give people freedom to an area where auditability is important. Let me just quickly run through it. What we do want to give our staff is our ability to look at data that was previously siloed, put it next to each other, compare it and really find new insights. So kind of your heuristic data exploration. That's a world of where you want to give freedom, you want to give where AI and ML play a big role. And that of course in the end after it will feed into project work because nobody's doing this for a hobby. You're also working as part of a project, a well-proposal or field development. And that is where you get into the middle pillar of this picture. And this is where we're very keen also at always the project workspace or project work environment will play a big role. The area where the vast majority of the work happens and where teams can cooperate. And maybe in that middle pillar you would see all your individual work in progress attempts at your seismic interpretation number 365, your simulation run number 2017. That is where the project workspace in my view will play. And then finally you move to the right because the end of the day we're about making decisions. We're about making megabucks decisions on the basis of the right data. And this is where auditability becomes more important. So maybe what we call, we're moving a little bit more from the technical process side to the business process side. And this is where it becomes important to record out of all those hundreds of simulation runs. These are the assumptions that underpin my low, mid and high case so that when you start drilling you will always see what assumption did you not get quite right. But that is where plan versus actual lives. And we see also in this whole ecosystem an important role for what we call workflow orchestration. And whilst we're working in this ecosystem we're very keen that always you will keep it safe also. That always you will make sure that we use the data only for those roles that we're allowed to, entitlements and obligations. A lot of data is available to us under a contract and we need to make sure that we don't overstep it. And that can all be done automatically. And also archiving, it was certainly when we started working on non-records disposal. This was a big thing because a lot of people said, oh my God, I have to go through all my data. Something like non-records disposal will become with OSDU should become a triviality. Once you're out of that middle layer where your project has come to an end and you put in your investment proposal perhaps you don't need to keep all your 800 runs that you didn't use and you get rid of it and it can all be automated with OSDU. Right, so how to get there. So the first step is of course loading data and this is really only the first step. And this is what we're at the moment working with and I'm not going to go into specifics because many companies here are going on the same path. We're in a mode at the moment where we're trying to load data to offer visualization first of all. And let's be clear, this is just the data, on-prem data is still right now the master. So this is the moving the data for the first time to the cloud. And this is not easy. I can tell you some people in this room have the scars for it already because of course it depends really on what your starting point is. How clean and how well organized is your data in the first place. So we're taking tons of legacy data to the cloud at the moment and there's all kinds of questions pop up and I try to put some of them on this slide. For example you get all kinds of dilemmas. So if I move the data to the cloud, do I clean it? Do I do a QC check before I migrate it? Or is it actually smarter to migrate it maybe a little bit in its current state and automate the upgrading of the data. And then of course as I said entitlements obligations is you really put your face in it so you really say you can wonder do I want to fix that again before I migrate? And then it's the actual transport of the data and what we call them pipelines, they're really hundreds of pieces of software that will take the data from on-prem to cloud. And of course speed is an important factor in doing this mainly with lots of log files but you can imagine there will be challenges when you talk about for example pre-stack seismic data. And then it's out there in the cloud and then you have a question do I need to do a back synchronization or not? Would rather not because that's a very complex business but of course not every company or not every part of your company or organization moves to the cloud at the exact same time they still want your organization to be able to work together. So there's all kinds of dilemmas and we're very keen to share our experience and to learn from you on this topic. Moving data is of course the first step and it's really nice to just if you think about photography using that as an analogy because when everybody took their pictures maybe by now Brown or so these photographs they did have no metadata with them. Nothing on camera settings or what have you. And of course photography has revolutionized with not only migrating photos to the cloud but also what you can do with the data is completely changed and that is what we're very keen on that this is not just a move to the cloud but it's really about changing the workflows really get people to work in a much more integrated way. So this is really what it's all about. In the near term what we're doing in Shell is really focusing initially on log data. We find that certain communities are extra eager for innovation in my experience often the petaphysis is under us a very numeric group of people they're very keen to get that data and for example see what they can do with AI ML. So that is something that we're doing but of course we had to start with the concept of a well in our industry. A well is an entity that moves through the organization from exploration to hydrocarbon accounting because it's often what I would call a coat hanger for a lot of the data because a lot of the data that we're loading is actually associated with a well. So that is the first thing that we've done so that truly a well is truly born in the cloud and that is quite a big job because you can imagine the concept of a well features in just about any application. So you're doing almost one of the most complex things on the front but we felt that that was really necessary. Then what we want to offer is really visualization for our staff so really visualization of data that was previously really siloed and it sounds very basic but there's already a lot of excitement for people to be able to see that. Seismic is a big thing. Seismic I would say both holds a maybe a majority of the promise of the cloud in terms of the ability through the cloud to speed up workflows and use compute for some seismic processing can still take typically nine months and you really want to flexibly apply that compute to that but it's also given the data volumes and the complexity one of the most challenging things. Not only data of course also applications moving to OSDU. We in our company we've got hundreds of them and that is a whole ballgame in itself whereas the moving the data is something that you learn you can you kind of you know becomes a little bit more repetitive and you can really speed that up. The applications that we have are of many shapes and sizes big and small many different companies so moving all that capability to the cloud is one of the things that we look at with all really in terms of a challenge but you've got to start somewhere and we hope to move the first what we call Workhorse application to OSDU this year. I marked a few things in red and one is the project workspace or project work environment where really we're looking for the OSDU forum to take the lead on the non-commercial part the non-competitive part of that because that refers back to that middle layer that I showed where we're very keen that we make good progress on that because we're moving everything to the cloud and you know you could argue whereas there's current complexity with on-prem and silo databases and maybe regional instances the potential is of course you make it even more complex unless you have some order in what you're doing in the cloud so some way of teams being able to really work together in the most basic way so that is to us it's very urgent and then another challenge that we see coming and again we hope to learn a lot about that what you're thinking about it is the reversal of mastership so as we're now loading data on-prem is still the master but you have to flip that around and that comes you can imagine with its own challenges how to do that and at the same time provide the business continuity because I don't know how it's in your organization but activity levels in our industry are very very high so we're moving all this to a community that is at the moment trying to keep all the balls in the air so it's not easy here we come to the team sport bit so what we try to kind of bucket activities and everybody has a role in this in this move to a different way of working really we're here for the OSD forum that's to me indeed appropriately we give that bucket A as far as we're considering anything that we can agree on is not competitive we're very keen for the OSD forum to lead that so that is the non-competitive part of the project work environment the non-competitive part of things like multi-region and I think the progress has been great we see great progress on the data type definition we got the CRS sorted out so really what we can with a very solid reference implementation is really what we're looking to the OSD forum for then of course the cloud service providers play a very key role multi-region and HBC will be some of the areas where they say this is where we feel that there's competitive space so they have a very clear role as well and then of course the software vendors and let's be clear OSDU is disruptive we wanted to be disruptive for a good reason but of course for the software vendors OSDU is also quite disruptive in terms of changing their commercial world what we are looking at from an IOC point of view is we're very keen that there is open competition through the standardization that OSDU offers really now a mini start-up company could compete with some cool software with the established names but established names of course they need to morph their business as well so it is an exciting space but the software vendors are of course key and then of course it comes to the right I can say that really we're keen as an IOC to have as little as possible in Bucket D we are in essence we are an energy company we're not a software company and there's no doubt that we will be moving to a kind of a pass or sauce environment over time there will be some bits that will be in-house the bits that we think are competitive and every company will have some of that but really we're keen to we don't want things to be bespoke for our own company because that actually defeats the whole philosophy of the OSDU forum and my final slide just wrapping up with this so to be clear nobody has done anything like this before we certainly have it we're certainly learning as we go it's good to stay humble we are definitely made some errors in our implementation everybody has we shouldn't be ashamed to say that and be able to we should be able to move to change course where we need to we're doing all this while the world is changing all around us technology is also at the same time moving at the same pace so you always have this dilemma do I now go with this or there's something else cool coming a little bit later but then you get the kind of the jam tomorrow dilemma where of course the world into this nirvana is an exciting one but it is a very heavy spent area and it's important to show some intermediate progress to our customers also along the way otherwise people start to lose the belief so you have to take it step by step and not gonna wanna go all the way right from the beginning there's some clear further enhancements of OSDU we all know them and lastly a plea and it's also very much in the philosophy that everybody can play at this we're very keen on the innovation marketplace we posted some stuff in there we got some great responses and we're following that up so that's maybe a final plea to really leverage the OSDU marketplace because I think it will speed up innovation and faster innovation means we give these superpowers to our people faster our workflows will improve and will make better and faster decisions and that is good for an industry that we're all in that is still essential to me a force for good and important for this planet thank you