Main content

On the right track - how a UK tech accelerator is bringing railways into the AI Age

Chris Middleton Profile picture for user cmiddleton May 31, 2024
We hear from rail operators about how a joint technology accelerator is helping an old industry onto an express route to the future.

Image of a railway

The UK is at the center of some major challenges in rail transport. On the one hand, it is the prime mover of railways and railroads - an innovation with roots in the 17th Century, and the invention of the steam locomotive in the 19th Century. 

The massive expansion of the railways from 1825 up to the beginning of the First World War saw Britain benefit from the economic growth, mobility, and connectivity that resulted from modern, faster transport links. New towns, communities, businesses, and supply chains sprung up around them - a phenomenon repeated in the US, of course, with a quarter of a million miles of railroad added before the war. Prior to the golden age of the motor car and the aeroplane, rail was an engine of US industrialization and growth.

On the other hand, however, the timing of those innovations left rail’s progenitor, the UK, with an ageing network that is difficult and expensive to upgrade without massive disruption - a challenge exacerbated by the difficulties of building new rail links in a small, densely populated country. The soaring costs, political controversy, commercial risk, and partial scrapping of the planned HS2 (High Speed Two) railway are a testament to that.

But there is some good news. The rise of new technologies such as artificial intelligence (AI), data analytics, sensors, the Internet of Things, drones, and more, allows rail operators to look at their assets, businesses, and customer relationships in new ways, unlocking the value of existing investments and helping to maintain them long into the future. 

But how to speed new technologies into a UK rail system that has, since the mid-1990s, fragmented and returned to its private-operator roots? (Any future Labour government would consider renationalizing the system in place of the 28 separate companies that run Britain’s railways today.)

This question has led to the creation of a technology accelerator for the rail industry - or at least, for an alliance of operators. Future Labs is the creation of four private rail companies - LNER, TransPennine Express, Northern, and Southeastern, driven in part by Ross Welham, Lead Digital Research & Innovation Manager for LNER.

Future Labs’ aim is to run a series of three-month pitching programmes - one a year - in which technology innovators from anywhere in the world can propose new technology projects to the four operators. In return, they get mentoring, expert advice, and the chance to work with rail companies to explore a concept’s fit and potential in the real world. 

Welham tells me:

Future Labs is so much more than just an accelerator in the traditional sense. We're not looking to take equity in companies that pitch to us, like some accelerators do. Instead, we immerse those companies in the industry and in our businesses - which again, accelerators traditionally don't do. Or they aren't able to do it, because they're not always affiliated to specific industries or organizations. 

So, this is an extremely high-value programme for companies that are looking to get into the rail market, or which want to develop and test something that maybe they haven't had the opportunity to do yet.

In this sense, Future Labs is about hitching credibility to opportunity via an alliance of four rail operators - with the UK’s infrastructure owner and manager Network Rail also involved in these conversations, adding what Welham calls the “full scope of the industry”.

Common challenges

So, why have four companies joined forces? 

The answer is interesting, in the context of potential re-nationalization. LNER, TransPennine Express, Northern, and Southeastern are all wholly owned subsidiaries of DfT OHR Holdings Ltd – essentially, the British government’s Department for Transport. 

With a total of six railways now under state control via this arm’s-length company, the government has been the UK’s biggest rail operator for some years, partly due to the failure of some private franchises. So, any re-nationalization programme would not be the outrageous retrograde step that critics claim. Roughly one-quarter of rail franchises are already run by the government, covering large swathes of central, northern, and south-eastern Britain.

All this contributes to the “common challenges” that those operators face, suggests Welham. He adds:

For a number of years, LNER had run an innovation programme of its own. And we invited other operators, such as Northern and Network Rail, to take part in that, but not in any official capacity. But we quickly saw the benefits of that collaboration, not just for us but also for the kinds of solutions that we are able to test and develop.

We saw the value of joining those things together. Now, companies like Northern or TransPennine can scope and validate ideas for the problems they are trying to solve. And, from an owning group level [i.e. the government], there is a lot of drive to be more collaborative, working on shared challenges and workstreams. And for us [LNER], it touches on pretty much every aspect of our business.

Might the alliance, and the Future Labs initiative, expand at some point in the future? To more operators, perhaps, such as the others under the DfT’s control? Colin Kennedy is Head of Continuous Improvement & Efficiency at TransPennine Express. He says:

In the near term, we've got our hands full running the Future Labs programme as it stands. Our laser focus is on this year’s programme being a fantastic success. But, for sure, I think after that we would be very open to consider future expansion.

Quick wins and minimizing disruption 

So, what about the new technologies that Future Labs wants to explore and implement in partnership with the world’s wannabe transport innovators? 

LNER recently partnered with British AI-based equipment health-monitoring platform Amygda to explore a predictive maintenance solution, with a view to minimizing downtime for customers - an innovation that has been adopted by Northern.

Northern Railway has upgraded a fleet of 40 trains to become smart, data-centric devices, and is using AI to collect data from different points across the network - with a view to maintaining equipment health and uptime.

Welham explains:

The Amygda solution was essentially OEM agnostic, and could take data from anywhere. They came through a previous Future Labs programme - prior to this formalized agreement - and are working with Northern today. 

Northern saw the value in what they were doing, and have worked tirelessly to help Amygda get immersed in their business and unlock some value from them.

He adds:

It’s all about opportunities for maintenance optimization. So, of course that's a huge area not just for us, but also for Network Rail. And it's a big area that a lot of people are working on throughout the industry.

So, with the interest of the UK’s national infrastructure owner in the frame too - not just individual operators - might this type of scheme be adopted nationally, if Future Labs can demonstrate a real return on investment?

Welham says:

Yeah, absolutely. I can't speak for Network Rail, but I do know that they are playing in this space already and have been for some time. 

They have assets in the billions [of pounds]. So, across the UK, their first task was to actually make those intelligent, or have some sort of intelligence in them. Then it’s all about what can you do with the data once you've got it. And that’s where we, with Future Labs, can really play a big part.

Good news - for all concerned, including the passengers, perhaps. But innovations aside, LNER’s Welham acknowledges that, in itself, the accelerator concept has limitations:

We are slightly constrained by the timescales of the programme, which only runs for 12 weeks [a year]. And so, we want to be able to demonstrate something in those 12 weeks. So, there are limits to what we can feasibly achieve within that timeframe. 

That said, we want a broad portfolio approach to the programme. We want high-value quick wins that can demonstrate the value of Future Labs, while simultaneously having a keen eye on the future. We are looking at potential opportunities that may not be deliverable in the short term, but could have a really high impact in the long term, and high value for the industry.

At present, not every idea pitched via Future Labs will be adopted across every operator, necessarily. While under overall, arm’s length control, each remains its own entity and has its own challenges and opportunities. As a result, some pitches or ideas will be a better fit for one company than another. 

Take AI. TPE’s Kennedy says:

We are still in our infancy at TransPennine in terms of the AI conversation. And that is partly why I'm so passionate about getting the company involved in Future Labs. 

But in all honesty, we are playing catch-up – a chatbot solution for our website, for example. But LNER already have a more mature version of that than we do. But from our perspective, there is a lot of excitement about seeing what comes through the accelerator’s pipeline.

He adds:

Collectively, between the four operators, we have agreed on four challenge areas for Future Labs. Those are: performance and operational excellence; enhancing the customer experience; developing people and talent; and a wildcard category.

In this sense, the four companies want to improve their services and operations, but are open to being surprised by ideas at the bleeding edge - innovations that may come from anywhere on the planet. Welham says that they are currently exploring proposals from as far afield as the US and Australia, as well as from UK talent.

He explains:

As you can imagine, there are loads that feature AI at present! But for me, it's all about how far AI is being pushed. Is it just the version of AI that everyone's got in their locker at the moment? Or are we seeing versions that are really pushing the boundaries and bleeding-edge technologies, which are definitely more the aspects that we are interested in. 

In terms of other proposals, there is stuff around gamification, safety, and security, with diversity and inclusion in there as well. And as you can imagine, one of the key challenges for everyone is disruption and how we get information to customers - in the right way, or in the way that they want. So, we have had a couple of proposals come in from those areas too.

Kennedy adds:

From a TransPennine perspective, the last couple of points that Ross mentioned are really hot-button topics for us, around monthly disruption and passenger communications.

At this point, it is worth stressing that disruption has two meanings in a conversation full of rail metaphors about “tunnel vision” and “going on technology journeys”. For organizations such as rail operators, tech disruption is something desirable and strategic. But for commuters, the other kind of disruption - to their journeys - should be avoided at all costs. So, an implicit challenge of Future Labs is ensuring that one type of disruption does not trigger the other, but helps minimize the risk or, at least, mitigate its effects.

In either sense, Future Labs has a critical role to play. Welham says:

Traditional accelerators go nowhere near the levels that we do through Future Labs, in terms of the benefits to all our businesses – the corporates that are running it and the companies that are part of it. 

And we have iterated it ourselves based on our learnings in this industry - and in this country. We haven't just looked at what others are doing and said, ‘They did it like this, so we should do that too.’ We've done what works for us, and what's right for us as a UK industry, rather than just copy-cat someone else.

My take

As befits the country where this whole industry started, back in the first Industrial Revolution. Now we are in the Fourth – which is about intelligence in the network itself – Future Labs would seem to be an ideal model for the future. 

On the right track, you might say.

A grey colored placeholder image