SDSW23: Foundational geospatial data infrastructure of Great Britain

tl;dr

The geospatial data landscape is undergoing a significant transformation, driven by the need to integrate data into wider systems and ensure its accessibility and trustworthiness.

The role of geospatial data in facilitating sharing across different systems and supporting national foundational infrastructure is crucial for the future, especially as new technologies like quantum computing and smart dust emerge.

Ordnance Survey is focused on enhancing data quality, testing, and governance to ensure that its geospatial data remains fit for today and tomorrow, while also emphasizing the importance of data accessibility, trust, and value.

Transcript

So good morning, everyone. Again, really pleased to be here. And I thought this morning, speaking to the 1Spatial team, I’ll just reflect a little bit about maybe a little bit more about the inside of OS and put us on a same footing as most of the rest of this room in terms of being on a data journey in our own organisation and the I suppose generational overhaul, if you like, of life inside OS and why we're going through those steps at the moment. And I think I said in the panel earlier on in that regard, it feels like we're going through a journey that I suspect most other organisations in this room are going through.

There's a couple of things which have maybe just draw out at the start. And one, thank you for the comment earlier on Seb about the naming, but whilst I didn't quite design master map and be there at the naming, I was certainly there at the introduction of it. So when I joined OS round about 20 years ago and over that period we've gone through quite a number of pretty big programs that are relevant, at a national level, the introduction of master map from previous generations, national addressing infrastructures for those of you who were around at the time, certainly 1Spatial were important too at the time with the introduction of Positional Access Improvement for some of you who shudder in the room at the remembering of such things.

But we've gone through significant data transformations, technology transformations kind of regularly, I guess, during the time I've been at OS, but the era we’re in right now feels possibly more significant than at any other time in the time life, certainly, being at OS, and it really matters in terms of the impact that national foundational data has to a nation. It really matters, and the conversation we're having with customers who probably over the last 10 years, 20 years or so have grown huge systems of their own asset data. And we're now rapidly moving into a generation where we want to share that data between each other. We want to have systems of data sharing, data exchange. We want to be able to connect insights across those different systems. And we're trying to understand our role, I guess, within that system and to create the right environment that makes that possible.

I always kind of like to use this slide and apologies to those of you who have seen me use it before. Because I try to kind of remind myself of our purpose every single day. And it's kind of encapsulated in many respects around those sort of 10 super use cases that our data underpins every single day today. And I use it internally just on a regular basis to remind everyone that what we do really matters. And you look around some of those kind of use cases, I suspect many people in the room will contribute to at least one, if not some of those. But for us we recognise our role in underpinning all of those use cases every single day. Some of those matter to, you know, the service delivery of government and the private sector, some of those matter in terms of making better policy decisions and policy environments. Some of those are expressly around improving the efficiency of the private sector environment and the way in which they're engaging with us as citizens.

But we're really conscious of the fact that our data underpins a significant economic value, whether that's inside organisations or to society at large. And in order for that to happen, it's important that our users can trust our data and place trust in the decisions they make. And I think, you know, Clare's introduction this morning really resonated with me in terms of some of that language. I'm really conscious that today our data underpins critical national infrastructure. I’d go as far to say, Ordnance Survey today, data today is critical infrastructure. It’s critical in underpinning the infrastructure that many of our users create uniquely or semi uniquely even on a global basis, underpins the land registration system of this country, underpins a trillion pound land and property sector, which is really significant. It has a history of underpinning applications like Streetworks that we’ll hear about a bit today and the underground infrastructure of Great Britain in many respects, you know, has a close relationship with Ordnance Survey data over the years.

Today, and of this year Ordnance Survey finds itself as part of government inside the Department of Science, Innovation and Technology. It's a new home for us. It feels a really great home to be part of in terms of the many other organisations with a sort of common association to data, if you like, to also sit inside that department. That department is bringing together data and associated technologies to really push the economic footprint of Britain forward in terms of innovation, in terms of technology, and it's great to see our contribution directly as being part of that. I think it probably also draws on the ecosystem that our data, our partners, our users collectively contribute to. I think it's also really important to be deeply engaged as part of the opportunity that we have all been given to have a policy function specifically championing and representing geospatial data, geospatial technologies and geospatial people for that matter, inside the heart of government.

The Geospatial Commission has just recently published its 2030 strategy. We're deeply I suppose connected, and we feel emotionally connected to the three missions that are laid out within that strategy. I think that's really important for pretty much everyone in this room to kind of be familiar with and to consider the contribution that you make your organisation makes to that bigger system that collectively makes Great Britain makes UK a more effective geospatial nation. On the slide, there are in fact three images, only two of which appear not to have come through, but something it was great to be on a panel earlier on and share that with Olive Powell. Olive and ourselves, ONS and Ordnance Survey have been doing a lot over the last couple of years now to really bring geospatial and statistical data foundations together to make sure they are greater than the sum of the parts that Olive was drawing attention to earlier on and making sure that both of those things together deliver more than they're otherwise sums. It's really important to recognise the work that is happening across government just now as part of the wider data framework, the wider data strategy that’s been launched and for us to consider our role withinside that so geospatial data, it's easy to say well environment is really important.

It's much more hard to kind of go into mainstream data strategy environments and say, don't forget about the geospatial bit, it's maybe the most fundamental. So we're doing a lot of work skilling across government in conjunction with ONS you know, kind of giving the analysts perhaps the more mainstream data analysts that sit across and giving them the confidence, if not the toolset to use geospatial data, more importantly, and to think about the way in which our data directly contributes to the wider integrated data systems and platforms that ONS are leading on across government at the moment.

The other thing and it was great to see GOBI earlier on in the audience and just our other significant contribution, which I think is sometimes overlooked within our own organisation as much as anything else, is the importance of geospatial standards. I'm going to talk about data sharing, I'm going to talk about systems of interoperability. You know, they don't just happen on their own and a commitment to thinking about the standards environments that we rely on to make this possible is something which our organisation continues to lead on and is fundamental. When we talk about foundational data, foundational data only works when it's interoperable, when it only works across the system, and our organisations commitment to both national standards here in the UK, IST/36, lead by British Standards Organisation, our commitment to ISO and TC/211 at international level. And of course OGC is really important and I can see a few folk in the audience today who are very much part of that community.

I think over the last 20 years I said we've kind of worked in, you know, against use cases, against applications. Yes, there's been a few paradigm shifts out there in technology terms. You know, it's always kind of, you know, humbling in one regard. But to think back 20 years ago as to what didn't exist in sort of technology terms, that just, you know, we take as a matter of course today the kind of the blue dot, if you like, on your mobile phone to always position you in the right place. You know, I hate to remind you all, it wasn't there 20 years ago, at least not ubiquitously like we use it today. And when you start to consider the environment in which our data is now being used and perhaps more importantly being expected to be used, the world is a very different place and technologically that's on a level of acceleration that feels both exciting but also quite daunting in equal measure.

So we started the day off today talking about data paradigms, you know, systems of data governance, systems of data improvement, thinking about whether our data was fit for purpose. And you swing around and look at an environment like that with all of the different technological advances, different use case environments that are ahead of us, and you start to ask some serious questions about fitness for purpose and what we do next. We can’t claim to everything we haven't got the capacity, the bandwidth, the knowledge, the skills to do a thing. So there's partnership considerations, but more than anything else is a prioritisation as to, what is our role as a national mapping organisation as a location data authority in underpinning this environment to work effectively.

Now, there's an element of deep geospatial skills have always been at the heart of our organisation. Transformations have always been at the heart of the organisation. I kind of think back, you know, we've got some wonderful, if you come down to Explorer House, some wonderful graphics as you come in through the front door of kind of, Ordnance Survey over the decades and in some case centuries and the periods of change and innovation and leadership that we've put in over the years to creating foundational data. At the core of our organisation, it was always built broadly around three core systems of operation in one way or another. You know, we captured data, we managed that and in the middle somewhere we derived a lot of value from that data, often in ways that it wasn't originally intended to do. And then we published that data.

So if you like the system of records, the system of management system of reference and the systems of customer engagement on the side and this morning I kind of already committed this sort of heretic exact moment of going when on the right hand side it was about producing a map a very tightly controlled specification that we designed internally, and we captured our data against it, we managed our data against it, we quality assured against it. You know, you can get to the point where you go, those days were easier. I don't say they were easy, but they were certainly easier.

Today, the vast array of data that's available to us that we're working with, that's data that we both collect ourselves that some of our suppliers work to collect for us, that we partner to bring on to our own systems. We're working with a far greater variety of data and data that we might not immediately collect ourself. But we know that you're collecting and that you’re provisioning and then expecting to be able to incorporate it into our data. Our systems of data derivation change dramatically. We don't know what queries, what expectations, what products and services that we're going to create from our data tomorrow and the day after. And so the flexibility that we're having to put into our systems of data management and the controls that we're having to put around that are much more complex, if you like, than we've built in the past.

And finally, on the on the right hand side of the journey, if you like, we've heard very clearly from our users that they do want less than the sum, they do want the parts, you know, the whole is great master map is wonderful, but actually within the database that supports master mapping or other products, you know, our users do want more granular access to that, more flexibility around the points of access into the data that sits behind it. And they've been encouraging us to work through that. However, we've thought really hard about this at a strategic level and apologies because at one level this is quite sort of an internally focused slide to our business and the way in which we work. But 1Spatial team are trying to encourage us to open some of that up today. And once we've got some great systems of data capture and management sources, we increasingly see our role towards the middle on the right hand side of this slide.

So I think there's a real challenge for us as an organisation which is predominantly, you know, century's worth of data capture excellence, if you like, surveying in its crudest form. Today to say yes, absolutely, we need to understand how to control those data inputs, but actually systems of data aggregation, the management of intelligence in our data, questions around validation, quality control, ensuring that we've got the right governance, that it's not just governance for our own data assets, but governance, that we can provide insurance to the market, around our data is something which becomes really important, that plays to some of our core strengths as an organisation, but nonetheless are things that we have to dial up if you like, in terms of our overall data landscape.

And on the right hand side for sure, let's think about the data, the products, the services that we offer and consider those to be quite different perhaps from how they were in the past. You know, I think we are now very much towards the right hand side of this slide. And, you know, I suspect if we did a straw poll in the room, we'd all come up with a different set of nouns, adjectives and verbs as to how to describe the right hand side of the data landscape to which we're working. But, you know, conceptually, over the next few years, we envisage working towards some form of digital twin of Great Britain. We see the systems that are beginning to emerge, the multitude of different digital twins that are emerging or systems of record that are emerging and thinking about how we act to facilitate the bringing of those together, absolutely feels really important.

I just wanted to kind of counter through kind of how we've then made some of that available.So the National Geographic database is something which, you know, I think hopefully most people in the room have a sort of some sort of affinity with. It's the database that traditionally underpins master map and all things that go with that. And whilst it's really complex in itself over the last few years, enabled by collaborations with the public sector, enabled by funding through the PSG, the Public Sector Geospatial Agreement that we maintain with government, we've already introduced around 70 new data types, data assets into the National Geographic database. They were features, descriptions, attribution of features which didn't previously exist in inverted commas “some sort of master map world” but are now very much part of the NGV. They're there to enrich the data that we hold today, but also start to facilitate some of the questions and some of the insights that our users are expecting of us. And this is something which, you know, we expect to continue doing significantly over the next ten years.

The long list of data developments, if you like, are only really constrained by the budgets that are available to go about doing that. But there's an incredibly long, rich list of attribution and particulars of the features which we collect from our data which will continue to deliver. And that data is now I said earlier on, it's kind of been liberalised in many ways. We're making that data available through a range of different APIs, both at title and feature level and giving our customers the ability to create their own menu cards of features that they want to draw on from the National Geospatial Database and publish that the way that OS select and build if you haven't already used it.

And so our transformation journey today is really kind of for me in the last sort of 20 years has been one of the biggest paradigm shifts in the way in which we're enabling our users to access the geography of Great Britain in a more flexible manner. It's never been possible to come and just say, “Can I come into your database” in a way, and extract from that and whilst that on one level is very liberalising in terms of access to data, as I said earlier on, it's putting a significant, I suppose, expectation, that we're feeling the pressure within that to ensure our systems of governance around that are adequate.

And I'm conscious of time and just, you know, I wanted to kind of get across the sense that for us it wasn't just about data, the expertise that comes in our own people and developing that. The network that exists across our partners, our users. You know, today we're sitting comfortably working with 5000 user organisations across government within 2000 different partners and developers that are working with our data. We have millions of users of our OS maps and the apps that goes with that and of course that range of discovery solutions and services that are built on that same data. So it's not just about the data so much as everything else that goes with that to do.

A couple of slides just to kind of finish off really to kind of give you an idea of the complexity and the simplicity that's kind of going into some of our back end at the moment. I would say, you know, you've matured as a geospatial organisation when you can move your system acronyms from three letter acronyms to four letter acronyms, as we're now in a world where the TLA is no more inside OS. But you know, our production platforms are now optimised to work far more flexibly with a broad range of different input types that we can assure on the point of input. Our core data services platform is our system of data derivation, where we hold data in one place and we can apply a range of different algorithms to that to extract value from it in ways which are far more flexible than we previously been able to operate. And the OS data hub, which is probably the bit that maybe most familiar in this room as the access point to our data.

It's a system of customer engagement its the place where select and build comes alive and we work collaboratively with our customers within that. I said earlier on that you know data was only part of it, but it is the fundamental aspect of our operation today. And we're going through a new journey of enterprise data improvement and our Enterprise Data Foundation model, which we expect to run for the next few years. With some humility earlier on, I looked around the room and I can see many other organisations that could probably use exactly the same slide. And whilst data governance is fundamental and we recognise inside our organisation the different nature of data governance in which we're working today, there are three fundamental building blocks to a multi-million-pound multi-year investment in our own data that we're working through.

Data testing and data quality is something which we're very attentive to today. Our data is being used in ways that wasn't previously expected. Computers do say no, so if data goes out in the wrong format with the with some wrong formatting anywhere else, it will not work. And we're really attentive to improving some aspects of our data quality and data testing around that. We've got generations worth of data improvement still being worked through a huge long catalogue from our own data scientists and cartographers who know only too well some aspects of their data that needs to be improved in order to meet the needs of the next gen applications. And our data development programme very much working in conjunction with a kind of overall public task and the customer requirements that we're gathering both from public and private sector to ensure that the data that we provisioned today at a foundational level for Great Britain remains fit for today, but also fit for tomorrow.

And I just thought I’d finish off actually, I suspect this is a slide which we'll see several times today, if there were three things that sit at the heart of our data principles that we try and live every day within our organisation, I sort of share these with you, I guess, and it's is kind of trust. Trust feels really important. You know, we we feel the burden, if you like, of our customers expecting to be able to trust our data. And that's a quite a subjective measure at times. You know, it's easy to put measures around quality metrics and so on, but we want to be able to stand behind the notion that our customers can rely on and trust our data, so we try to bake that in.

It's really important that our users can find our data to discover it, to find ability and accessibility to our data is really important. So are we cataloguing our data right, are we publishing metadata to go with that, do you all have confidence in the data that we're making available? And then finally, you know, for data to be valued, it has to be used. And there's a real sort of emphasis on ensuring that we are connected to the impact that our data is having to the use cases that our data is underpinning and learning more and more about the way our data is being used in practice and those same data principles as well. So they apply to the sort of foundational and public task aspects of our work today, they also transcend into all of our products and services, be those, you know, commercially or public sector oriented. The same principles, I would suggest apply to any of our work from a data infrastructure perspective.

So I'll leave it there. Thank you for the opportunity to contribute to today's proceedings and enjoy the rest of the day.

Yeah, you ok for some questions?

Yeah. Yeah.

John Hartshorn: Does anybody have any questions for David? You’re a remarkably quiet audience. You might be the quietest audience I’ve had in 33 years, there’s one there. Do you have a loud voice or do you need the microphone?

Jaymie Croucher, Transport for London: David, it’s fantastic to see the road map to left development. What's the single biggest challenge you think that faces geospatial data over the next five years? That's 2030?

David Henderson: Wow. I mean, and you saying that in the round? Not specifically necessarily for us, but I think for geospatial data generally, it's making sure it finds its way into the wider systems of data that are being developed across at pace at the moment. I think I was really careful in saying that I didn't sort of, you know, try and make geospatial out to be special in some way and have our moment or any of those sorts of things. But I think, you know, there is something that's really important the next few years and you know, systems of data are happening at pace at the moment. There's a lot of existing data in amongst, you know, this type of community that needs to find its way into some of those new systems and to recognise that, I can't remember the great terms that were being used earlier on. I think Simon might have had a couple which were kind of, you know, whilst, you know, kind of spatial data might not be special in that kind of context, it is really important.

And, and I think, you know, we risk overlooking the importance of getting geospatial data embedded in some of those new systems of governance, whether that's your private sector, whether that's public sector. And for me, the single biggest reason then to make that, you know, make that a reality is when we start talking about sharing across systems, more often than not, it's location, which is the question which demands that data come across that. And if we don't have that kind of common sort of glue enabled by geospatial, these systems will not come together and we will not have that kind of national foundation that we probably enjoy more today.

Thank you.

John Hartshorn: The thing I was going to ask you, but you might not know the answer to because the graphics were brilliant. I love them, but what is smart dust?

What a smart dust. I can guess, but I think I might be wrong. Yeah, I'm going to duck it, I'm afraid. I think there's probably people in this room. They'll have a better answer on that than me.

I'm looking forward to hearing what that is. And the other one I wrote down interesting was quantum sensing, yeah and quantum computing.

David Henderson: I mean, just the the I mean, and actually I think that is really important for, you know, just looking forward over the next few years, if, you know, if we think we can produce data at scale today and I think we can process things at scale and if you think about space and, you know, the amount of information that's coming to us today with quantum computing thrown into that, you know, the speed at which decisions or data can be liberated and decisions can be made.You know, I think if we ever thought about us working in a real time sense, I think we're getting into a paradigm where the expectations and the computing power will be there to enable it.

John Hartshorn: Yeah, brilliant, brilliant. Thank you very much, David. Thank you.