Transforming Railways with IoT

Alex Jablokow
Illustration: © IoT For All

Some of the most transformative IoT applications come in industries that seem part of the previous industrial revolution or even pre-industrial. One such example is freight railways. With 140,000 miles of track, the US freight rail network is one of the largest and widely regarded as the world’s best freight rail system.

Seven large Class One railroads and over 600 other railroads carry many goods, half of it bulk cargo like coal, oil, agricultural products, metals, and construction materials, and half shipping containers, what is called intermodal traffic. It is an essential part of the economy.

Difficulty Maintaining Railway Assets

Freight railroads build and maintain their own infrastructure, and in fact, most passenger railroads lease their track from freight railroads. So freight railroads have to monitor and repair all of their rolling stock and all of the track, signaling, and other infrastructure, a massive job.

Railroads spend 19 percent of revenues on maintaining and adding capacity—making that task easier and more efficient is worth a lot to them. Adding tracking is expensive (from $1 million to $2.5 million a mile for the existing right of way, much more if the land has to be acquired). Volume increases have to come from more efficient use of the existing track while maintaining safety.

Around 80 percent of the track issues that cause derailments result from vertical displacement, normally happening over time. This should be monitored continually, with incremental changes detected and noted, but there has been no good method. IoT provides several possible routes to the knowledge of track geometry. For example, Canadian National Railway is testing track inspection cars of various types, as well as accelerometers in locomotives that note any unusual track geometry or rail joint issues.

Managing Railway Assets: Even More Difficult

Railways have long been desperate to identify, locate, and track the condition of all of their complicated, expensive, and dangerous physical assets.  The prehistory of IoT is full of clever ways managers tried to understand the assets they are responsible for, and railways are no different. Their attempts included the multicolored proto-barcodes of KarTrak in the 1960s (the odd railway allergy to the letter C is clearly long-standing). After the memory of that failure faded and new technology was developed, RFID-based automatic equipment identification (AEI) using wayside detectors was introduced in the 1990s.

AEI provides ID and location for every car, decreasing cost and increasing sensor and communications capability means that every railway is looking at real-time monitoring of tracks, fixed assets, and the functionality of every part of locomotives and railcars via IoT.

A manager will be able to see that a change in current or a falling propane tank level indicates a snow melter at a remote switch will soon fail, and thus send a crew out with the right replacement parts long before the switch is frozen in the wrong position. But that incident will joint thousands like it, supporting an understanding of mechanical failure patterns across the rail network.

Transitioning from Data Scarcity to Oversupply

While sensors of various kinds can reveal a lot, it would be a mistake to rely solely on that data for making operational and maintenance decisions. Context is important. Other essential information comes from maintenance logs, both for rolling stock and fixed assets, and historical data like traffic, rolling-stock flows, and weather.

The most important thing to remember is that the data is acquired, analyzed, and used to pursue an understood business case. There is always a temptation to focus projects around data availability, which certainly raises the chances of success. Thus, personal recognition, whether or not that project actually furthers the organization’s goals. But focus on the business case is essential for long-term success.

Like many legacy heavy-metal industries, railways have a large supply of skilled staff with defined workflows and an intuitive sense of the equipment. If they are not included in planning and see no benefit, their resistance can doom even promising organizational initiatives.

Data from existing equipment recorders will be incorporated into data flow; existing low-bandwidth radio installations will continue to provide connectivity, heavy manual switches in remote locations, many decades old, will have basic sensors added without ceasing to require an engineer to climb out of the cab and throw the switch through muscle power.

Never Underestimate Heavy Metal

IoT is a transformational technology that mostly makes its changes slowly and incrementally. In the coming years, locomotives and railcars will seem pretty much the same as ever. Simultaneously, they function for years and decades longer because of condition-based maintenance, run faster over safer tracks, and carry increasing volumes of goods, all thanks to IoT.

Alex Jablokow
Alex Jablokow
Alex specializes in writing about interesting and underexplored niches of IoT (particularly industrial IoT), medical devices, and healthcare IT. Despite being an engineer and a science fiction writer, he inclines Luddite-lite. Along with the Eleph...
Alex specializes in writing about interesting and underexplored niches of IoT (particularly industrial IoT), medical devices, and healthcare IT. Despite being an engineer and a science fiction writer, he inclines Luddite-lite. Along with the Eleph...