Originally established as an arm of the British military in 1791, the Ordnance Survey (OS) is one of the country’s most venerable institutions, with its orange-backed fold-out maps an integral part of family camping trips for many.

But as smartphones and satellites have increasingly encroached on the OS’ traditional areas, the company has been hard at work establishing itself as a vital part of national infrastructure - one that could have ramifications for years to come.

This of course, entails greater interaction of AI, so I spoke to the Ordnance Survey to find out just what this transformation looks like.

AI automation

“This has been a technology business since day dot,” Nick Bolton, CEO at the Ordnance Survey, tells us, harking back to the early surveying days of theodolites and chains, where tools were constantly improved in order to out-do European competitors, “it’s always been technology that has enabled a better map.”

Bolton points out that the OS has long been focused on using computer technology to improve its process, starting the transformation in the 1970s with plotter and digitiser tools to scan paper maps - which then had to be printed in more advanced ways than ever before.

The modern version of the OS is far more technical, and Bolton, whose background was in computer vision, highlights that the organization has been using AI technology for over a decade.

This began with implementing automatic feature extraction from aerial images, as the OS scans the whole of the nation every three years - the type of high-resolution image data, Bolton notes, that is ideal to train a neural network on.

“We’re at the high end of the market,” he notes, “everyone thinks AI is associated with large language models and transformer networks - actually, we all know that AI has been around for a hell of a lot longer than that.”

“We see AI in the bracket of automation - more than a human replacement,” Bolton adds, noting that with Gen AI, having natural language conversations is “another obvious area we can apply AI to”.

Agentic AI is a major interest of the OS going forward, as it looks to offer its huge cache of knowledge and data to partners, both in the public and private sector.

“We’ve been getting questions about that for a very long time,” Bolton notes, “we really have to make sure that whenever we do anything with AI, we have to think of the things that have made us great us in the past - being authoritative, being assured, and being trustworthy - none of those things can go away in the AI age, in the same way that they were there in 1791.”

Bolton notes that secure access is another crucial consideration, and points to the OS’s AI Charter as a demonstration of how important this is to the organization. But AI usage is still gradual, he says, noting it is an “organic process”, pointing to the importance of trusting the decision made using OS data in situations such as warning against building new houses on a flood plain - there needs to be assurance that this decision can be traced back to concrete data.

“That there’s always going to require some form of human check on it, until such a time that we can feel confidence in the output,” Bolton says, “making sure we go slowly, preserving that this decision or that insight can be traced back to this set of data means therefore I can have a high confidence in it…it all requires a change in the way we think about how we produce our data and how we attribute our data.”

"Everything happens somewhere"

The OS data is being used in a huge range of use cases - with government agencies featuring alongside utility networks and construction firms.

Bolton notes that the OS’ goal is to try and facilitate as many answers as possible, highlighting that, “location is the obvious way to connect all data.”

“The term we use internally is that everything happens somewhere,” Bolton notes, “location data matters in a wide variety of situations.

“What we really have here is a general purpose technology, a bit like the semiconductor or the pneumatic tyre, it has such a wide variety of applications, we can’t be expected to own the whole of the stack, our task should be to stick to our mission, make the best data, and then work with others to incorporate that data into many other solutions.”

Geographical and location insights are “widely applicable…it’s universal,” Bolton adds, "everyone has got ‘where?’ questions…geography, location and engineering play to the sensibilities of Britons”.

“The great news is that from a GB perspective, we’ve got a tremendous resource in the Ordnance Survey from a data perspective, and we’ve got a real software capability in this country from vendors who understand that domain.”

“In that ecosystem, where you’ve got lots of people with lots of different types of ‘where?’ questions, we’re not going to serve all of those, but rather work with a series of partners, and the good news is that’s a world we’ve been building for a long time.”

Bolton gives the example of a homebuyer and a mortgage provider - the former will want to check OS data to make sure there isn’t a radio mast being built in their back garden, but the latter will want to ensure there are no issues around flooding or other hazards, so they can ensure they get a good return.

Its ownership by the UK government also means the OS works with national bodies, particularly with the Office for National Statistics and HM Land Registry, as well as the intriguingly-named National Underground Asset Register (NUAR), which monitors the UK’s network of underground pipes, cables and other sub-surface infrastructure.

Bolton notes that the built environment as a whole is already a mixture of focus from the public and private sector, highlighting how, “place by its nature, is a collaboration between private and public.”

This means that along with offering data to the likes of Santander and Sainsburys, and advising solar panel installation companies on which houses might be a good option for future rollouts, the OS has also been working alongside the Met Office to form the Government’s response to the recent wildfires, blending the former’s weather data with OS map information about the location of wooded areas which may be further fuel, or spot any critical infrastructure such as electricity substations which may be at risk.

"Special insights"

The OS has changed so much since its 18th Century beginnings, but as Bolton notes, it is still dedicated to democratising access to its “special insights” - albeit now changed from warfare, towards national infrastructure.

"If you look at our organisation, we look like a technology company,” Bolton notes, “but we didn’t always look like that - and that’s always the journey that you have to go on.”

Answering the need for such varied use cases will involve greater working with partners, but as Bolton says, the OS seal of approval will also look to provide data that is trusted and authoritative.

“We’re always seeking an ever increasingly efficient factory,” he adds, highlighting the split between AI product and AI productivity, “how do make sure we can bring the efficiency gains, and we’ve always done that, and achieved more, in order to not scale back what we do, but to scale up our ambition, and go and do more.”

“I don’t think that will ever change, in the same way that we provide wonderful guidebooks alongside maps, and a mobile app.”

Mike Moore is Deputy Editor at TechRadar Pro. He has worked as a B2B and B2C tech journalist for over a decade, including at one of the UK's leading national newspapers and fellow Future title ITProPortal. When he's not keeping track of all the latest enterprise and workplace trends, he can most likely be found watching, following or taking part in some kind of sport.