Meet Gareth Dennis, our new Transport Data Lead

Gareth Dennis joins us as a Transport Data Lead. In this blog he introduces himself and describes the challenges currently faced in the transport sector when it comes to data.

I’m very lucky.

As a (relatively) young railway design engineer, the opportunities I’ve got essentially by being mouthy about things I care about have been far wider than I could have imagined when I started my career in 2013.

And so it is with my joining Open Innovations as a transport data lead. I’ve always had a deep fascination with the opportunities that data can provide people who both make and live with transport policy, not just in terms of the big picture stuff, but also when it comes to fiddly, messy engineering. To be given a chance not only to espouse this to a new audience but also to be able to steer some of those solutions is incredibly exciting to me.

There’s no shortage of data that we already collect; indeed, we have data coming out of our ears… The challenge for end-users is in usefully tapping into it, and for data-wranglers it is in knowing what the end-users need. Hopefully I can provide a bit of a bridge on that front!

One example is the pressure that a changing climate is exerting on transport infrastructure. Railways in the UK face a triple-whammy: assets are degrading faster than they can be renewed, they are being hammered by more extremes of weather than ever before, and we are asking them to carry more trains at higher speeds than they have since their construction. It’s a battle that we are losing, as evidenced by the increasing number of weather-related failures we are seeing.

I’ve written elsewhere that the use of asset-related data is the key to turning this around – and I fully expect to visit this in greater detail for OI.

That’s the bottom-up view. Top-down, the questions are different. Is our railway network the shape it should be? Are we making the right decisions about transport investment? Do we actually know the real impacts of making changes to our transport network?

On my weekly podcast (#RailNatter goes out live on Wednesdays at, I have prodded at the edges of some of these questions, trying to use a data-driven approach to break free of some of the “well, the railway network looks like this” limitations that stymy much current policy.

This also feeds into one of my passions – the democratisation of transport policy making. Whether through more openness in policymaking, by devolving more power to the UK’s regions or through the proliferation of better tools for campaigners and journalists, public transport will increasingly fail to satisfy people’s needs if we continue on the current trajectory of more obfuscation, more behind-closed-doors strategizing and more reliance on shady (and often useless) business case modelling.

One of the questions I get asked a lot when I’m popping up as a talking head is “what does levelling up actually mean for transport” – to which my answer is invariably about frequency. Within the M25, most people don’t have to worry about timetables as they can turn up and hop on a train within minutes. For the majority of people outside of the M25, there’s no such luck.

My challenge is this: can we create a mapping layer that shows the maximum number of trains calling at any single platform of all stations on the UK rail network, per hour? One number per station, perhaps shown as a heatmap.

Why per-platform? Well, because this is a reasonably proxy for the number of trains going in each direction. Given that the majority of stations in the UK have two platforms for two tracks, this would neatly give the peak service in either direction.

On its own, the number of peak services at stations is journalistically interesting, but combined with train capacity and local population data it becomes a powerful tool for understanding potential (or, if you like, stymied) demand… At a recent Transport Select Committee evidence session, I compared a few stations to explain what levelling up needs to do.

Over the next six months, I am hoping to look at a variety of data-led network development tools with the OI team… Firstly, I'll be looking at applying some science to a framework me and some data colleagues have already been working on to establish optimised transport networks for populated areas of a variety of sizes. After that, I hope to kick off the development of tools to look at long-term public transport demand potential based not on black-box modelling but on assessing past performance of transport projects.

When I talk about democratisation of transport decisions, making those kinds of numbers easily accessible to local people, politicians and journalists must be priority number one. As my first challenge to the OI community, I hope it ends up being one of the most useful!