Network Rail has embarked on what it calls a "digital railway project", which will enable it to track and monitor every facet of the network through a vast array of connected devices, including trains equipped with lasers.
"We want to connect up the whole railway end-to-end, to understand where every piece of equipment is. We have some general idea where the trains are most of the time, but we want to know where every bolt, sleeper, tree is without having to send people out to them," Clayton Nash, Network Rail Telecom head of telecom product told Computing.
Network Rail commissioned a new network with an initial capacity of 1TB - with plans to increase to 8TB - which saw it move from PDH and SDH systems to a single a MPLS in an effort to base "the whole railway on an IP backbone".
The tendering process was meticulous with trials for each of the vendors shortlisted, because if something went wrong when the network was live, it'd be catastrophic.
"A key thing was safety resilience and whether or not we could replicate our existing PDH services over the network in a safe way. Because if a hiccup in the network means a signal gets missed, that has serious consequences," said Nash.
"So we conducted what was described by one of the respondents as the most comprehensive, technical trial they've had to participate in as part of a tendering process," he said.
"We spent six weeks running every scenario and at the end the Cisco solution passed with flying colours, so they were selected and it was good value for money."
A key benefit of the connected network, Nash explained, is that the remote monitoring it allows will reduce the amount of time engineering staff are exposed to potential danger. "Safety is a key aspect of what we do and one of the ways you make people safe is by not having them walk up and down the track, they can't be hit by a train if they're not there," he said.
With the new network, engineers are able to view potential faults via computers and tablets using laser line pattern recognition technology.
"Basically, you spray lasers out the front of the train and you get a 3D map of everything around you. At the same time you put some high resolution cameras under the train and you can photograph all the fasteners and sleepers," said Nash.
"Through this simple application of computing power, you can have somebody virtually walk down the track," he continued, adding this connectivity makes maintenance repairs more efficient.
"Ultimately, you'll need to send someone out to screw things down, but you want to do that because it needs to be done, not just in case. You'd also like to do it in advance of something going wrong, so when we can spot stuff in advance, people can go out and do preventative maintenance. If we know what the whole railway looks like, we can do it much more safely and efficiently."
Nash told Computing how Cisco has enabled the deployment of a trackside IP voice solution that enables better communication than aging and expensive copper cables did.
"As the power requirements for driving trains are changing, it's becoming increasingly difficult to run long lengths of copper for phone services up and down the trackside," he explained.
"Now we're able to deploy an IP voice solution along trackside. We're able to run that over much longer distances and meet the project timescales whereas we had no real option if we'd had to use copper, it would have been a horrifically expensive exercise."
The benefits of the new network could potentially be felt beyond Network Rail, according to Nash. "We're already speaking to rural broadband companies so they can benefit from the strange places our network goes where others have chosen not to go," said Nash, who described how ultimately, it's passengers who'll benefit from Cisco's input.
"Once we have an instrumented railway and understand exactly where the train is, how fast it's moving and whether there are any maintenance issues, the railway is going to run better because you can anticipate failures," he said, adding it'll enable better services.
"You can also better predict when the trains going to turn up, so no more of that 'the train's five minutes away, wait, it's already turned up' kind of stuff, it'll be a better form of train estimating," he concluded.