How ‘predict and prevent’ equipment failure is now keeping trains on track
A rail switch 250 kilometres from the nearest depot was about to fail. There could have been catastrophic danger at the worst, or, at the least, several hours of repair delays for the entire rail network.
Both risks were avoided because sensors in the vital ‘point machines’ that switch rails remotely reported irregular electrical signals from the motors to the network control centre. There, AI identified the potential failure that was then corrected by maintenance before it could falter.
Victor Borges, Thales Digital Product Manager for Predictive Maintenance in Ground Transportation activity says "In the UK, we have supported Network Rail in designing a system that supervises more than 40,000 connected devices in real time”.
The bottom line? Train travel that is safer, more reliable, and much more cost-effective for railway budgets.
“Rail operators spend more than 80,000 euros on maintenance for every kilometre of route”, Victor Borges calculates, “That’s a total of 15 to 25 billion euros a year. Imagine if they could save just one per cent of that a year? That would represent between 150 and 250 million euros of savings annually!”
"Imagine if rail operators could save just one per cent of maintenance a year? That would represent between 150 and 250 million euros of savings annually!” Victor Borges, Thales Digital Product Manager for Predictive Maintenance in Ground Transportation activity
Smart infrastructure flags problems before they happen
The Predictive Maintenance revolution in ground transport “is only just beginning”, he says, as the Internet of Things (IoT) on the rails is advancing connectivity and creating vast amounts of high-quality data.
“Thales is going through its own digital transformation and we are already adding sensors and IoT capabilities to all our new products, upgrading legacy equipment, and even applying it to third party systems.” Victor Borges explains, “Data coming from that smart infrastructure allows AI to spot potential problems before they happen. And the more data we collect, the smarter the AI becomes. It is a virtuous circle.”
The next step? Sensors on trains themselves will multiply many times the instantaneous and continuous information that is today coming primarily from stationary sensors on the tracks.
Borges says, “The Communications Based Train Control (CBTC) system—the driverless metro signalling system—itself a great example of Big Data – is producing a great variety of data every second. They tell us a lot about the health of the CBTC system and about important sub-systems such as communication. So, with AI and Machine Learning, you have real-time diagnoses for a constant health check of both train and track.”
"With AI and Machine Learning, you have real-time diagnoses for a constant health check of both train and track.” Victor Borges, Thales Digital Product Manager for Predictive Maintenance in Ground Transportation activity
An early warning system made possible through Thales technologies
Creating this distant early warning network for ground transport is natural for Thales, as Victor Borges notes, “We have the chain of technologies to make us the leader in rapid and efficient Predictive Maintenance for rail and other ground transport. In addition to our long experience and expertise in signalling and rail network management, we are able to put together an optimal solution for the most efficient use of the Internet of Things with both moving and stationary sensors.”
“We know how to continuously extract and secure the integrity of the data, move it instantaneously, and analyse it rapidly to identify potential problems before they become real failures. We have now real experience in Machine Learning to create the AI algorithms and then apply them to understand the meaning of the incoming data”.
Victor Borges adds, “The most recent companies to be integrated into Thales complete the technological supply chain required for transport Predictive Maintenance. Gemalto is playing a central role in veracity and security of networked data. Gemalto brings unparalleled command in hardware device security, as well as versatile IoT connectivity modules. And Guavus has leading capability for Big Data analytics”.
Sherlock Holmes on the rails?
So, will Predictive Maintenance mean that Thales is putting a legendary detective like Sherlock Holmes on the rails to find the maintenance risk before the equipment fails?
Victor Borges laughs, and is more measured in his analogy, “I would rather say that we are offering rail operators a reliable early warning system. Instead of having to send out train crews to determine what went wrong on the track and come back with the equipment to fix it, we can analyse in a few seconds a source of probable failure to allow the rail operator to establish what needs to be done. If anything, Sherlock was correct when he said: “When you have eliminated the impossible, whatever remains, however improbable, must be the truth.”;
So Thales is helping companies to eliminate the least likely problems to identify the most probable one”.
"We are offering rail operators a reliable early warning system" Victor Borges, Thales Digital Product Manager for Predictive Maintenance in Ground Transportation activity