Representatives from Elgin, DfT, Highways England, TomTom and Ordnance Survey gathered to debate ‘Data as National Infrastructure’ at the HighwaysUK event in Birmingham on 17 November 2016.
Summary highlights of the messages from each are given below, together with video links to their presentations.
Introduced by BBC Transport Correspondent Richard Westcott, the introductory audience poll revealed the biggest barrier to widespread use of autonomous vehicles on UK roads was standards and interoperability with infrastructure and other vehicles.
First to open the session and invited to put the challenge into context was Shane O’Neill, Chairman of Elgin. Elgin delivers the roadworks.org service that provides a single portal and a public reference source for live and planned road and traffic information on both the strategic and local road networks.
Data represents a particular challenge for the highways sector because of the lack of a sustainable model for sharing data across boundaries. Data is stored in disparate locations, in different supplier systems, in individual local authorities and strategic road authorities. Accessing and sharing this data is challenged by existing contractual, operational and cultural barriers.
Why is this? Is it really acceptable to swallow the party line that its all to difficult to deliver a single network across 200+ different organisations? Given the physical nature of what this industry does, you’d think interoperability across boundaries would be hard wired into our consciousness. But that’s not the case.
Could the problem be interoperability between different providers in the market? Not really, as other sectors are also fragmented, yet still manage to achieve a consistent approach. So fragmentation of an organisation is an excuse, it’s not a reason why data does not flow across our networks.
Is it that there are not enough standards? Well, the sector is blessed with a lot of standards, so this isn’t really a plausible reason.
Could it be money? But it seems there is no absence of this, and the sector is peculiarly fond of pilot projects. There are many of these where lots of money is spent to prove that technology is not an impediment to sharing data. As public sector and commercial organisations, we’re happy to provide our data for a pilot project, but what about an ongoing sustainable model for the exchange of data? That’s what’s missing in our industry.
We don’t have to look too far outside of highways, in public transport, to see how a sustainable model for sharing data across boundaries is possible, Take the pump-priming of the creation of Traveline to enable public transport data to flow. We now have real time scheduled information from lots of organisations free-flowing into our applications.
Data is hungry, for vast volumes of information, in real time, and we need to get our heads round the problem that data has to move across boundaries and be permissioned. Many organisations, including local and strategic highways authorities, authorities, logistics companies, satellite navigation companies are all part of this ecosystem. All this information needs to be recognised, permissioned and in real time. To do that, one needs to set up an enabling mechanism, and to do that, you need to understand all this conceptually.
And perhaps this doesn’t come particularly easily to our Highways sector civil engineering mindset, with a 10-25 year project view and focus on physical infrastructure, where IT (technology is the carrier for data) tends to be delegated to the IT department. Which is akin to handing the keys of the asylum to the inmates. Because data is not exactly the same as technology. And IT people also like their big projects, and big infrastructure investments.
Data needs to be separated from IT. It needs to be set free. which is not the same as it being free. It may be cheap to replicate, but it is expensive to create and maintain. Some of that which is funded can be free. If its not funded there may be charges. We have to be able to permission chargeable, non-chargeable, and sensitive information through a trading platform for data.
This is a concept familiar in lots of other value chains, but deeply unfamiliar, and I suspect uncongenial to the civil engineering culture that dominates highways.
Things are changing, and I’m really privileged to be followed by some speakers where a strategic approach is being taken to datasets, and increasing interoperability between strategic and local highways.
We are moving into a new stage, where we have to learn what happens in other value chains, and think data and information, as opposed to technology.
Next up, Nic Cary, Head of Digital Transformation at the Department for Transport, illustrates the business case for open data, particularly for local authorities.
I tend to view data as part of the bloodstream of our modern economy, so if it’s part of the bloodstream, why would you silo it? Why would you keep some cells from being nourished by it?
In my department, a lot of policy people say that data is too technical. The point is not to think about the technical aspects, but to understand the benefits.
The challenge of course, as experienced by colleagues in the Ordnance Survey and the Hydrographic Office, is that traditionally, mapping companies have made their money by selling an encapsulated form of data as a paper map, and people appreciate you have to pay for that.
In this rather scary new world, where a lot of data is available for free, it’s difficult to understand how to build a business when the raw material is free.
The point is, and it can’t happen overnight, but one has to move to a model where the value is what is added to turn data into information.
For highways this is incredibly important. There has already been much progress made opening up data within Highways England, following the pressure from Francis Maude, Minister to the Cabinet Office for public bodies to open up their data. In the commercial world, this is a much more challenging problem, because data has to be maintained, kept up to good quality etc.
But if you look at the impact of not opening up data, look no further than local authorities, who I really, really want to help and get more cohesion and coherence with.
Inside local authorities, typically data is siloed, and in highways that a real problem because it means you can’t actually see the maintenance problems yourself, and can’t understand the impact of one activity on another. By opening it all up, you can look across the entirely of the organisation, which is as true for local authorities as it is for a business.
Understandably, this feels challenging to many local authorities, facing continued budget pressures and priorities in Adult social care. But by releasing the data they gain greater coherence internally in their highways maintenance programmes.
They can also support the greater good, as opening up this data helps people who are journey planning get a better picture of when they should travel. They can also understand with greater ease which routes will actually work, which ones are busy, because at the moment we are gradually being dominated and taken towards a business model that is largely controlled from California. and whilst I have a love and respect for many of the companies in that field, there is an alternative, which is to use open tools, open data, and do it for ourselves, and make the whole of our economy flow better.
Tony Malone, Chief Information Officer, and Executive Director for IT at Highways England shares his vision for a publishing / subscription platform to access Highways England data, and why this is so important.
In my first 18 months in the job, I came across excellent projects being delivered internally to maintain and manage our networks:
- The National Traffic Information Service is an awesome project, but only has an internal facing staff of around 350 people, a web interface, and data available to download if you know how.
- We have 30,000+ active devices on our network, and we use floating vehicle data from mobile users. But we bring it in to do our job, we don’t communicate it out very well to the public.
- We don’t integrate with the Urban Traffic Monitoring Control (UTMC) data that local authorities use. Why not? We have the technology to do that so there are no barriers to doing it.
- Asset data is a major issue for us, and we keep our estate in pretty good condition, obviously. But we store all that data, all the LiDAR data, terabytes of it. I know what a pipe looks like, for example, so why don’t we store the exceptions to that, then communicate that back down to the supply chain so they can build better things in the future.
- We have a system that reports on all our technology. It tells us every fault on all our technology in real time. What do we do with it? We maintain our kit. What we really should be doing is feeding that back into the supply chain so they can better their products, so we don’t have these failures in the future.
- We run the 5th largest fibre network in the country. It’s an awesome bit of infrastructure and there is lots of capacity there we can use for commercial benefit and there’s no issue’s about doing that. We can do that and really want to work with our partners, in both highways and other parts of the business to say ‘do you want to use that network, because we’ll enable it for you’. there has to be commercial terms of course, but the fact is we need to think differently and we’ve been very much about using data to maintain our road network, rather than using data for the greater good of the country.
So, my vision, when it comes to data, is you should be able to subscribe to any of the data that we hold. It’s publicly available, but we need to be freer with it and actually advertise how we do that with you. Likewise we want people to publish data to us that they think is important. In particular, local authorities. We publish using standard protocols, so people can interact with it, but it’s often too difficult/overly complex to work with. We can make that easier for them.
I really want the car to tell me and user what they should do. Not an app that you have to go into and put your route in, it should know what you do because you drive the same routes most days. If you go off on holiday, or other route, then perhaps use a journey planner, but the cars can do this sort of thing and I want to see that happen.
I think I’ve joined Highways England where we are thinking data and we’re thinking big. We’re working with some awesome companies to help us deliver that and we want to be open. My message here is that if there’s things you think we can do better, we’re here to help.
Douglas Gilmour, Key Account Manager, UK and Ireland for Tom Tom, represents a private business perspective view into the debate, Tom Tom is an example of someone who collects data and represents the user of that infrastructure; road users, drivers, car-makers.
We all know that data is big. We all know that data presents lots of fantastic opportunities for the transport sector. And it also presents some massive, massive challenges in terms of how we utilise or exploit that data. The exciting this about this session and breaking down some of the barriers between different parts of our infrastructure and those managing it, is also to be able to capture more and more information from the user and understand consumer satisfaction, as it were, in our transport sector.
There are lots of analogies / similarities between the public transport and road sectors. But there is also a clear separation between data use, ownership, management, and maintenance, much like road maintenance, management and transport sectors. I believe we need to be very careful how we think about the data ecosystem.
There are real responsibilities on transport authorities to maintain and secure data. Before that, we need to recognise that data is an asset. An asset to each authority, and each company and supplier into the road maintenance and management sector. It has to be managed, secured, stored, processed and understood in order to make use of it.
My contentious proposition is that there is a real battle for control in the data world at the moment. There is a battle for your data as a user, as driver. There’s a battle for your data as a mobile phone user and someone who browses on websites / commercial supplier sites.
There’s a real battle for control and influence of road users by authorities. Whether that’s using data from fixed infrastructure, or data from floating car sources like Tom Tom, or from other publicly available sources, it’s a real battle to capture that, maintain it, ensure its suitable and fit for purpose, and then to use that for the benefit of the public.
Data is just like any other national infrastructure, in that the public interest is at stake. Not just safety, but the public interest in terms of how we use data to serve road users and public transport users. My proposition is that he who has, and uses, and exploits the data most effectively and efficiently, wins, and by win we need to define that in terms of the public interest.
And finally, Neil Ackroyd, Chief Operating Officer, Ordnance Survey, make the case for prioritising the data we need, rather than the data we already have.
I spent the first three-quarters of my career in the West Coast, in the 1980’s, building a technology company that brought you GPS. At the time, we really had no view on how successful or disruptive it would be. We just saw a really cool technology that we could do something with. It’s taken 30- years for that technology to drive a whole host of questions that we’re now trying to answer. Questions around how do we deal with this new technology that delivers new capability, whether it’s connected autonomous vehicles, internet of things, smart cities. So I’m more interested not in the data we have, but the data we need.
We’ve done a very good job in the last few years working closely with the Department for Transport making sure the data we have can be brought together into a single view, that can start to address some of the concerns that Shane O’Neill from Elgin raised. A dataset that gives us a consistent view of what vehicles can travel, what the road widths are, what the network design topology is, and who maintains it.
But that’s not the data we need.
The data we need needs to be designed for the future we aspire to have. The data we need is data that can mediate between an autonomous vehicle and the general public, between an autonomous vehicle and society.
The map we need in the future is a machine to machine map, it’s not a map for people to consume. At the moment we are collecting data, and there are projects that government are looking at across the world as to what does that look like? What are the standards we require to allow autonomy in our infrastructure? What does a map look like in 2025? A map that a car consumes or a vehicle consumes, or vehicle maps itself. How does that get mediated? Who takes responsibility and authority for that mediation? Are we going to allow every car to make a judgement about the software it runs, or are we going to provide an infrastructure and some level of authority for that to happen?
Autonomous tractors have been available for several years. Quite recently, it’s now possible to purchase these, with no legislation to control their use. this is an interesting example of technology finding a way. We think we can control it, but actually technology finds a way. We have similar issues with autonomous cars and drones, we have regulation and design issues to deal with the technology that’s disrupted us, that we’ve let disrupt us. So looking forward, the question I’ll be asking is what is the data we need for the future? We need to better use the data we have, absolutely, that’s a given. But let’s design the data we need, because otherwise the technology will get there before our ability to deploy and use that technology will. For Ordnance Survey, that’s the real challenge.
The Highways sector is an industry that needs to disrupt itself. The ecosystem is still dominated by a number of large data organisations, whether that be west coast, or system integrators. We have to accept its a very complex, mixed ecosystem. There are many questions and many answers. Open data is definitely part of it. But so is significant investment in new data, data as a infrastructure.
When we talk about investing in new physical infrastructure, we really need to be talking about investing in new data infrastructure. These conversations are starting to happen, in groups who are starting to think about what data in 2025 might look like. But I do have a concern, that we’ll get there before the questions are answered. For me there is an urgency about our industry pulling together, and not just trying to solve the problems in our rear view mirror, because they’re are real and we know they’re there. But we actually have to solve the problems looking forward 5 years and 10 years for any infrastructure investment.