Next gen railway operations intelligence

Introduction: network fragility and reputational risk  

For as long as there has been a rail sector, service disruption (for example, delayed and canceled trains) has been a major cause of customer upset and significant reputational damage. We have all experienced the frustration of a delayed or canceled train – and some of us have gone on to resent the rail operator for it.  

Sadly, such disruptions are common. This is because the railway network is an integrated fragile system. For example, one delay can have a domino effect on the network. Indeed, one incident can lead to a full day of disturbed traffic and many upset passengers. Addressing those issues requires an improved traffic management system.

The first requirement is to provide accurate, tailored information to the customer. Punctuality is a key driver of business performance, both for passengers and for freight. Your customer may accept a delay, but it is becoming increasingly unacceptable not to know why there is a delay, or to provide this customer with erroneous or incomplete information.

What are the main challenges to overcome?

In pursuit of better traffic management (and customer experience), one of the main challenges is to precisely manage network problems when (and preferably before) they occur.

This entails anticipation, alert management and resolution. However, the systems we have today are not built around such problem solving. In the past, the main role of operators was to trace routes and timetables – but today, we must rethink this approach. In particular, we must take into account the digitization of the operational process, moving from only managing disturbances once they occur to a more proactive approach of prediction, consideration, prevention and anticipation. This is because, as, an old saying goes, ‘prevention is better than cure’ – complex problems are easier to solve if they don’t happen in the first place.

However, it is unrealistic to assume that all problems can be prevented, at least with the technology we have today. Take traffic jams as an example – these can be created if routes are poorly planned. Operators who make this mistake often suffer for months (as do their customers). Even today, with all our advancements in technology, it is still very difficult to maintain a robust, achievable timetable. So, in the inevitable case of a disruption, operators must be able to quickly perform short re-routing and re-planning. This translates into an understanding of the situation of the network, with a real-time tracking system.

Being able to adapt the transport plan effectively to cope with eventualities requires a clear understanding of real time needs – but also of future requirements. As alluded to previously, the ability to anticipate needs in advance is highly desirable, particularly for complex problems. For example, the risk of disruption increases when the network is forced to work in a degraded mode (eg. a signal failure). It also increases during periods of exceptional demand – for example, an event like a football match must be planned for and anticipated, otherwise insufficient trains will be available and passengers left dissatisfied.

Planning ahead: technology to optimize traffic management

Such forethought is becoming more realistic. Indeed, with systems becoming increasingly intelligent, it is possible to track and anticipate the arrival of passengers. This is a major aid in organizing future transport plans. As networks are interdependent, big data and data analytics can help to solve problems – for example, by correlating and analyzing data in real-time from several network assets – turning this data into relevant insights.

For instance, an issue with a train in Strasbourg can delay another train in Marseille, which could lead to a total disruption of the transport plan. This is an extremely complex problem that requires significant human resources to predict or solve (assuming that it hasn’t escalated so far as to be unsolvable). Anticipating traffic impacts and preparing solutions is, therefore, difficult. As a solution, AI technologies can help us evaluate all possible future scenarios and combinations – predicting anticipated issues and offering potential solutions to the operator.

Decisions on train priority could be made based on the number of passengers inconvenienced. Trains can be deployed to meet demand, based on detailed individual data on passengers’ scheduled movements. This approach is based on global decision support and the implementation of automation to manage workload and make complex decisions. Anticipating connections between two trains may also be needed, to reduce the impact of delays on passengers.

Back to our signal failure example; some of these digital solutions can help with the management of degraded modes, for example, using a digital twin of the network to simulate the effects of the degraded mode, or to check in advance if the network is enough robust to absorb extra traffic.

Growing initiatives for a centralized network control

Today, operators are putting a number of these technology solutions into practice, particularly the introduction of greater automation, thanks to Operation Control Centers (OCCs). Thanks to automation, the geographical remit of each of these centers has increased.

For example, in France, the current objective is to replace 1500 legacy signaling centers with 16 Rail Operating Centers (ROCs) that use new digital communication systems. Here, AI, IoT and data analytics could improve the system further, for instance, through video image analysis and macro data, which allow better supervision. As a result, the passenger load could be monitored on the network, to automatically adapt the transport plan in near real time.

Some barriers to digital transformation

Although we have made some progress with automation, the integration of all these new technologies is easier said than done. A minor change can disrupt everything.

For instance, introducing new modern trains without modernizing the underlying infrastructure won’t improve the system’s overall performance. For another example, adding driverless trains offers increased operational flexibility, but this kind of change requires both physical assets to be upgraded and, potentially, a complete update of the transport map. Nevertheless, metro systems are moving in this direction.

Therefore, successful digital transformation requires us to take a holistic view of all the parts of the railway network, one that considers integration.

Back to our command center example. Further improving the efficiency of these centers is partly an integration challenge. Within such a facility, all supervisory equipment may be provided by different specialized companies, not all of which are compatible with each other. Instead of having all functions contained within a single system, this requires the operator to move from one isolated application to another. As a result, operations are nowhere near as effective as they could be. We know this can be done better; we propose an ecosystem approach…

Conclusion: go further with partner ecosystems

To deliver more flexible deployments and provide efficient solutions tailored specifically to operator needs, operators should choose a broad and highly experienced ecosystem integrator that understands their goals and challenges.

As a leading global Intelligent Rail transformation partner, Capgemini has worked for years alongside rail operators across the world as an integrator, supporting their transformation to next-gen traffic management. We can provide your organization with the software ‘bricks’ that will help you to build a robust, failure resistant and user-centered transport system. Want to find out how we can help you?…