Transforming the rail network with digital innovation | I-CIO
The UK’s rail network is at full stretch. David Waboso, group managing director of Digital Railway, describes how digitalization will unlock capacity, boost performance and enhance passenger journeys.
Almost every business, in every industry is facing both the threat and opportunity of digitalization. In many cases executive teams are already turning their organizations towards new business models, new rhythms of innovation or even new markets. But some are wrestling with intractable fundamentals that demand large-scale renewal before they see a big digital dividend.
The world’s oldest passenger rail network was initiated in the UK almost 200 years ago, and today its 16,000 kilometers of track, signalling and switching infrastructure is being pushed to its limits. Over the past 20 years, passenger journeys have more than doubled to 1.7 billion a year, with a further billion annual journeys expected by 2030.
Some of that rising demand will be met by the addition of new network infrastructure such as London’s Crossrail (currently Europe’s largest engineering project) and High Speed 2 between London and northern England (with a budget almost four times that of Crossrail). But faced with such acute pressures, the UK’s Digital Railway program, an industry-wide initiative led by state-owned rail infrastructure manager Network Rail, is looking to digital innovation to unlock capacity and maximize performance across existing activities and structures.
That is no minor challenge when the infrastructure ranges from state-of-the-art train control and traffic management systems to the “dumb infrastructure of trackside semaphore signals, coloured lights on sticks and Bletchley Park-like relays,” says David Waboso, group managing director of Digital Railway.
His organization is charged with accelerating a vision for the digitalization of the rail network based on three pillars: the rollout of ever smarter trains, the creation of intelligent rail infrastructure and the use of command and control systems for managing the network.
Digital remedy
“Because of its characteristics, rail is a hugely expensive infrastructure,” says Waboso. “The challenge is to get the most performance and capacity out of what we’ve already got in the UK. Traditionally that might have involved longer trains and platforms or the re-siting of signals. But we’ve now largely exhausted those approaches, so we need a different way of unlocking capacity, and that is where digital comes in.”
At the heart of the existing network is a rail signalling and rail system that is only crudely aware of what is running on it. “In essence you have a dumb infrastructure. It knows it’s got a train on it but it doesn’t know much else — what kind of train, how fast it’s going and so on,” says Waboso.
“If you can start to computerize both the train and the infrastructure, with a centralized command and control system overseeing that, you can really squeeze more performance and capacity out of the system. With that you also get greater safety and reliability, conflict avoidance on the network, valuable information to customers, drivers and staff, much better asset management and a more sustainable model because your intelligent assets can predict their maintenance needs and reduce the cost and disruption from renewals.”
Pioneering moves
The idea of computerizing the train and track and having a centralised command and control system is hardly new. Many metro systems around the world have applied that model, as well as several mainline railway lines in mainland Europe and the Far East. In fact, as capital programmes director at London Underground, Waboso led the regeneration of London’s Victoria Line, where digital systems have pushed the frequency of trains to 36 per hour.
“The Victoria Line has shown what can be achieved with digital technology without investing in expanding the network [though the program did also include the renewal of ironwork, track, foundations signalling and power]. That achieved about a 30% overall capacity increase without a single kilometer of track being added,” he says.
Now the first mainline rail schemes in the UK are starting to follow suit. In the past year the Thameslink network has deployed automatic train operation using the European Train Control System (ETCS), a central command and control system that oversees and optimizes the movement and spacing of trains. Thameslink has also attained a 30% increase in capacity at peak times with trains now running every 2.5 minutes at rush hours on its routes through central London.
Can such a formula have the same results across the much wider network infrastructure? The next phase of the Digital Railway program runs from 2019 to 2024, with the aim of digitizing a “very significant proportion” of the millions of conventional signals across the network, says Waboso, with highly digitized trains running on that.
Together with the UK’s Department of Transport and the industry’s Rail Delivery Group, Digital Railway is helping to guide the specification for these trains. “In the coming years, we will not only have a bedrock of digital railway infrastructure, but an increasing number of trains fitted with all kinds of digital technology. It’s taken a while for that generation to become the norm but there has now been an explosion in the number of trains with huge amounts of data and processing power, positioning technology, passenger loading, performance and so on,” says Waboso.
Coupling those to a highly intelligent traffic management command and control allows the optimization of train movement through conflict points such as busy junctions or terminal stations, he says: “Then when something goes wrong (whether it’s an asset failure, a passenger problem or whatever) the system can re-plan everything rapidly and get it flowing again — something that would challenge human capabilities.”
Targets and inspirationWaboso is reluctant to predict just how much capacity digitalization can squeeze out of the UK’s network, but he is clearly looking at the 30% gains that have been achieved elsewhere as a reasonable target — with some caveats. “It’s wrong to put a blanket number on it because each line has a different mix of trains, different characteristics, a different number of stations, different line-speed profiles and different terminal stations,” he says.
But what does that look like from the customer’s perspective? By 2024, if plans come to fruition, he says, 70% to 75% of passenger journeys will be on some kind of digital railway. For the customer that means more frequent, more reliable trains, less overcrowding at peak times and better informed and connected journeys.
To get there, Digital Railway is drawing on the digitalization experience of other industries, especially the airline sector. And that extends to how it works with a wider group of digital technology specialists.
“One thing we’ve taken from airline industry is Early Contractor Involvement (ECI),” says Waboso. “Since joining in 2016 I have tried to re-energise, to change the paradigm of our relationship with the supply chain.”
Traditionally that relationship in areas such as maintenance has been too focused on “fix-on-failure,” he says. “We design it, tell suppliers what to build, and they put it in under our direction. This is one of the few industries where suppliers have put in kit and have no ongoing relationship with it. What we want to do is move to predict-and-prevent on maintenance and to buy outcomes. If those elements are put together, we will benefit from a different relationship with the supply chain, and an incentivized whole-of-life mentality around procurement.”
That means Digital Railway is looking at a more collaborative, longer-term model, with shared risk and shared incentives. “I want suppliers to be financially incentivized on the performance of their equipment. Not just when we first turn it on but through whole-of-life,” he says. To the extent Digital Railway might choose to start buying outcomes from suppliers, based on metrics such as availability, number of train paths per hour, or percentage points of improvement. “We want to buy a number from suppliers that says, ‘if you don’t achieve that then you don’t get paid as much; and if you achieve more, you get paid more.’”
The fresh thinking has also meant new invitations to the discussion table. In its most recent phase, ECI Phase 2, Digital Railway has extended the net of early contractor involvement to include many non-traditional suppliers.
“There are all kinds of associations with rail-sector contracting — high barriers of entry, slow to innovate, risk averse, very standards-based and prescriptive. But we want to stir things up a bit and see what might be possible,” says Waboso. “We want to encourage innovators to enter this industry, to bring in people who are real disruptors: companies like Fujitsu, McLaren, and Panasonic, as well as some of the much smaller innovators.”
There is a clear desire for a co-creation model, he says. “We are looking at a much more collaborative and longer-term model. We definitely want outcome-based contracts and leave the supply chain as free as possible to innovate within those.”
He likes how suppliers are already more proactive. “The supply chain is starting to be bold now, challenging us by saying, ‘We think you should do it this way.’ The challenge for us is how we take this forward.”
Digital twinFulfilling such ambitions will inevitably mean suspending certain sections of the network during implementation. But Waboso has plans to leverage digitalization to minimize disruption. “One of the points we emphasize in our business case is access: how can you do this digital stuff without closing down the railway.” His group is working on the concept of a ‘digital twin’ that creates a simulation of all the physical entities on the network.
“That will allow us to simulate all our work before undertaking it in the physical environment. If this is a 99.999% analog of the physical infrastructure then we can test problems and designs. We can mimic a planned update 100 times, test different permutations, simulate failures and see the consequences before we do it for real,” he says.
That points to the prospect of embedding even more machine intelligence in systems. “As we operate new traffic management systems and see how the railway responds to various problems we can write algorithms for how we want that to optimize the response. So it enables a continuous learning process,” says Waboso, suggesting that rail should be as enthusiastic about artificial intelligence and machine learning as any other branch of transportation.
That underscores the pace at which the UK rail network now seeks to embrace digital transformation. Before Network Rail decided to accelerate its Digital Railway plans, its target date for a fully digitized network was a distant 2062. By resetting that to the new horizon of 2029, it has signalled that even a 200 year-old infrastructure can be rapidly disrupted by digital.