Trains with Brains: How can AI and Digital Transformation improve the railway experience for consumers?

Optimised maintenance

ProRail, which is responsible for carrying out maintenance on the Dutch national railway network, has begun to roll out just such a network of IoT sensors across its assets in recent years. These sensors don’t just monitor the condition of tracks and signals, though. Instead, they are connected to a predictive maintenance system which uses AI to forecast when equipment is about to fail, based on its performance and history and on algorithmic modelling.

This just-in-time maintenance model minimizes downtime and reduces the need for disruptive manual inspection – hence boosting efficiency and enabling better use of equipment and resources. ProRail’s system has reportedly driven a 25% decrease in failures (covering sign-posts, switches, overpasses and detection systems); furthermore, the technology has been credited with achieving a 3% increase in the number of trains running on time.

ProRail has even set about incorporating passengers’ phones into its sensor network. This comes via a customer app which records when the train goes over a bump. If the firm receives enough reports from a single location on the line then it will be flagged, and workers will be sent to carry out inspections and take further action, as required.

Belgium’s state-owned rail company Infrabel, has set out on a similarly large-scale digitalisation project. Under the ‘Smart Railway’ program, Infrabel has implemented a predictive analytics infrastructure, which incorporates data points such as the visibility of sign-posts, changes in power consumption and possible overheating in trains' shaft sleeves. This information is compiled and then integrated into a single, all-encompassing data pool, spanning Infrabel’s network.

And the company has some stiff challenges: passenger numbers have increased by 50% since the year 2000; high-performance trains are putting increasing strain on tracks; and a busier schedule has meant that maintenance is more disruptive and hence harder to coordinate. Fortunately, digital innovations can assist on all fronts here, enabling tighter scheduling and a higher standard of maintenance and repair – and ensuring safety for both passengers and workers.

Another development that is augmenting asset inspection is the use of autonomous drones. These can be set on pre-planned flight paths, in order to monitor tracks and other assets over long distances, working quickly, efficiently and with minimal human intervention. Automation can also transform the role of train drivers. Thameslink, for example, has self-driving trains which can increase the capacity of congested routes. Nevertheless, they retain human drivers who check platforms and close the doors before leaving stations.

Conclusion

Digital transformation is certainly no small thing – but it holds huge potential and can enable operators to make vastly better use of both their assets and their resources. What’s more it provides a granular view of the network that it would have been simply impossible for rail firms to have obtained in the past.

“Digitalisation really does contribute towards improving the effectiveness and efficiency of maintenance and thus to a safer, more reliable and better-quality rail network,” commented, Luc Vansteenkiste, Infrabel’s Director-General of Asset Management, writing in Global Railway Review.

Innovation may be daunting – but it presents an enormous opportunity for change.