Operators of rail networks have been managing and maintaining their assets since railcars were first invented. This process has undergone significant evolution over the past two centuries.
Asset maintenance can be categorised in a number of ways. Let’s simplify with the example of a light globe:
In the very early days of the rail industry, maintenance generally followed a reactive or ‘break-fix’ pattern – when an asset broke, it was fixed. It didn’t take long for rail operators to attempt the creation of a maintenance schedule that would allow them to maintain their equipment before it failed, especially for assets where failure could be both costly and dangerous. This saw the birth of preventative and predictive maintenance hybrids in the rail industry.
*While any type of maintenance performed to prevent asset failure can technically be called preventative maintenance (PM), PM usually refers to maintenance performed at intervals pre-determined as optimal.
These can include:
This type of maintenance is most appropriate for situations where “abrasive, erosive, or corrosive wear takes place, material properties change due to fatigue, embrittlement, etc. and/or a clear correlation between age and functional reliability exists”. Source: wbdg.org
Centuries of rail asset management have seen us move beyond using the reactive break-fix method to scheduled or preventative maintenance, then make use of modern technology to perform condition-based maintenance and progress all the way towards predictive maintenance.
Even without sophisticated data tracking, maintenance teams began to identify patterns around when equipment was failing and create interval-based maintenance schedules with some degree of reasoning behind it. This could be scheduled by time, or later by mileage, but the maintenance had little to do with the actual wear and tear of each individual asset. Operators simply knew how often things broke and tried to schedule maintenance before this happened.
Not long after scheduled maintenance was born, an early form of condition-based maintenance emerged. This involved a daily visual check of things like wheels and axles for cracks or other damage. One practice was painting a white mark over the wheel tyre and hub so that any movement was immediately noticeable with the naked eye. Another was to strike wheels with a hammer to listen for differences in the ringing to detect flaws.
This manual inspection of axle integrity was complemented in the 1930s by more advanced electrical and magnetic particle tests to help identify cracks. By the 1950s, early forms of ultrasonic testing were used on rolling stock axles to more accurately discover faults so repairs could be performed before disaster struck.
Computer advances in the 1990s increased our ability to track more data than ever before and gave us a new ability to monitor a wider range of precursors of asset failure such as excessive vibrations, unusual temperatures and so on.
This meant that rather than simply performing maintenance on an item because the schedule said so, the asset could be monitored more closely and repaired when it began to show signs indicating that failure was imminent. The widespread adoption of computers by corporations also saw attempts to move away from paper records to spreadsheets and early enterprise-scale software solutions.
This ‘condition monitoring’ paved the way for intelligent predictive maintenance. Rail operators now have the option of fitting equipment with technology that allows it to send data about its condition straight to a central system without any manual intervention.
Technology has advanced so much over the last two decades it is now possible to track numerous aspects of your assets. In theory, you could run the numbers and calculate the likelihood of when every single lightbulb in every station and carriage is most likely to fail – but do you really need to?
Yes, if the globe is the only one lighting the way for a vital process, but this is rarely the case. For non-critical assets whose failure does not have serious consequences or impact (service-related or otherwise), reactive maintenance is still appropriate. The cost of unnecessary preventative replacements or installing sensors onto everything for predictive maintenance can be greater than the cost of simply running the item to failure and reactively repairing it.
For other critical infrastructure or rolling stock though, allowing them to fail before replacing them is simply not an option. This is where new inventions like intelligent vehicles, the Industrial Internet of Things (IIoT) and advanced sensors have opened up a whole new way to manage assets. There is no denying that when used appropriately, predictive maintenance has been a game-changer for rail operators.
Complex operations like rail transport carry high risk. With so many moving parts to manage, rail operators must make use of highly-evolved, cutting-edge asset management technologies if they want to avoid critical, costly failures.
As we all know, failure of assets in the rail industry is often more significant than a burnt-out lamp at a station. In these situations, predictive maintenance really is the holy grail of asset management.
Predictive maintenance for the rail industry represents an exciting point where maintenance and even engineering teams can use real-time information and big data analytics to glean accurate and reliable intelligence on how assets will perform in the future. Sensors can be installed on various components and modern on-board and wayside systems come with their own diagnostic and health monitoring systems increasingly often – all opening the door for transmitting performance/condition data in real time.
This goes far beyond looking for cracks in equipment. Essentially, it’s like looking into a crystal ball and seeing – with a high degree of accuracy – what will happen to an asset unless something different is done in the management process.
In addition to maintenance methods for meeting regulatory requirements, sensible operators of complex assets have come to realise that a combination of reactive, preventative, condition-based and predictive asset management yields the best outcomes. Rather than dogmatically clinging to one method, they strive to find the optimum mix of maintenance practices to maximise asset reliability while minimising costs. This approach has been gaining popularity in recent years and is called Reliability-Centered Maintenance or RCM.
RCM is not a new concept. It has been used successfully in other fields, but adoption in the rail industry is not commonplace. However, initial research by the Mineta Transportation Institute at a heavy rail transit agency in North America indicates that RCM has a positive impact on rolling stock availability and reliability – it is a practice worthy of further investigation and evaluation by any savvy rail asset owner.
Ultimately, the goal of all public transport players is to ensure their services run safely and effectively. The right mix of maintenance practices, especially utilising modern technology like predictive maintenance enablers where suitable, will allow rail businesses to handle issues appropriately before, as or after they arise.
The benefits are clear: passengers enjoy reductions in service failures, while operators can continue to kick KPI goals by consistently running services on time. Infrastructure owners are spared catastrophic failures and will extend the useful life of their assets with the most appropriate maintenance.
RCM has in the past been viewed as a white elephant due to its relatively frequent failure compared to other practices at the implementation stage; not because it is philosophically unsound, but because organisations were unable to successfully execute it as a maintenance strategy without buy-in from senior and middle management and commitment from the entire organisation.
Making the move to RCM may not be as impossible as it may seem. Rail operators around the world have already begun to use EAM software as a tool to effectively manage their assets and store maintenance data. Having an easily accepted EAM system that is capable of accommodating each type of maintenance method and every asset type could be the key to a smoother adoption.
For example, Trapeze’s EAM system was tailor-made for rail from the ground up and can be used for all types of rail assets and maintenance methods. Rail agencies that use it have seen excellent results in terms of staff adoption and training: one of our customers, Sound Transit, showed an adoption rate of nearly 100% and time savings due to only needing to train new employees in one system.
The varying assets that make up a rail system – rolling stock, infrastructure and facilities – require industry-specific approaches and practices to achieve optimum maintenance. The right EAM software will provide operators with a clear picture of all asset types system-wide and where each item sits on a maintenance plan, whether that be condition-based, predictive, preventative or even reactive.
Your best bet to accomplish this without having to pay for extensive customisations is with a rail-specific solution.
 International Railway Journal, Railways and risk; can prognostics reshape asset management? (2016), http://www.railjournal.com/index.php/rolling-stock/railways-and-risk-can-prognostics-reshape-asset-management.html
 The Railway Technical Website, Train maintenance (2017), http://www.railway-technical.com/trains/train-maintenance/
 Reliability Centered Maintenance: A Case Study of Railway Transit Maintenance to Achieve Optimal Performance (2010), http://transweb.sjsu.edu/MTIportal/research/publications/documents/2913_10-06.pdf
Enterprise Asset Management