What is Asset Condition Monitoring (ACM)? A simple enough question, one would assume, with an equally simple answer. Perhaps not. A quick check online throws up a myriad of results. Even refining the search to include ‘railway’ still presents a bewildering array of options.
This article will attempt to answer that question. However, with its finite limits, it will be able to touch on some, but will by no means cover all, potential aspects of ACM.
For the purposes of this editorial, we will go with the following definition: “A system that allows a user or maintainer of an asset to be made aware of that asset’s condition for one or more monitored parameters of that asset.”
One example could be the state, time to transfer between states and motor current required to move a passenger door. So there is an asset – the door including motor and command and control equipment – plus state and time taken to transition between its states of open and closed. The ACM is therefore the monitoring system that measures and reports on those aspects of the asset.
That seems all fairly simple, but now the fun begins.
Traditionally, a maintainer of an asset would have operated one or a combination of ACM processes that might be classified as ‘Find and Fix’ – wait for an asset to stop working, find the failed component and fix it. This may be cost effective in some industries but, given the severity of delay attribution penalties in the UK, probably isn’t a wise choice for widespread adoption.
An alternative is to adopt the OEM’s advice on periodic maintenance. This is probably presented as a conservative schedule. After all, the OEM doesn’t want its equipment tarred with the ‘fails in service’ brush and so it probably recommends asset maintenance or replacement whilst there is still remaining ‘useful’ life in the asset, which will be wasted.
Either way this is not really ACM at all. Waiting for an asset to fail and respond rather than monitoring its condition, or replacing items periodically because that’s what the maintenance manual calls for, take no real consideration of the asset’s condition at or prior to the act of maintenance. These are purely reactive systems, acting in response to a fault occurring or the arrival of a periodic maintenance window, with no understanding or consideration of the actual condition of each asset. One option results in unpredictable disruption when faults arise or are wasteful of remaining asset life due to the periodic nature of some maintenance.
Crafting a solution
The aim must be to use technology to monitor the asset’s condition, or at least a specific number of key physical parameters relating to the condition of the asset – examples might include motor current, relay room temperature or track geometry, depending on the actual asset under consideration. Then take this acquired data and build a database of the true condition and operation of individual assets.
As the old saying goes, “knowledge is power”. So, to analyse and extract the information from large quantities of points monitoring data, Balfour Beatty has spent fifteen years creating and refining a software solution called AssetVIEWTM. This system makes use of the scalable processing capabilities available in the rapidly evolving field of cloud technologies to identify and report alerts from the data in real-time, provide the ability to drill down into the detail of these alerts for the purpose of root cause analysis, and provide early indication of impending failures.
The efficacy of such systems is heavily reliant upon the quality of parameter selection, rules and thresholds, which Balfour Beatty has developed and evolved in conjunction with its domain specialists. As a real- time online solution, the AssetVIEW system is accessible 24/7 from wherever maintainers need it. With the ubiquitous use of smartphones and tablets, it can exploit Push Technology: a subscription model whereby users register (or subscribe) to alerts for specific assets. Similar in concept to the alerts received from social media applications, these give real-time information about events of significant interest. It is as if your point machine has a TwitterTM account!
These modern techniques are exactly what Balfour Beatty has provided for SMRT (Singapore Mass Rapid Transit) where the company’s Points Condition Monitoring system (PCMALERT) has been chosen and installed to monitor 150 point ends over 32 locations. It is responsible for acquiring real-time condition and event data, providing SMRT with the decision support required to optimise the work of maintaining a network of point assets, review asset performance and create improved maintenance schedules. This is particularly important as the points were all installed within a very short timeframe and require maintenance at much the same time, a significantly risk-laden and highly challenging undertaking.
Now this approach is all well and good for discrete assets that are, in effect, static or only occupy a limited area, but for something more linear, such as rail, a slightly different approach has proven to be more effective. Rather than just deploying trackside monitoring of particularly problematic locations, track maintainers can team up with train operators to have unattended monitoring devices fitted to passenger service vehicles that automatically record the characteristics of the track over which the vehicle is travelling.
Then, in the case of TrueTrakTM, these devices utilise either 3G or Wi-Fi hotspots to download the acquired data to a central repository for analysis and presentation in DataMapTM.
Once again, this system was developed in conjunction with domain specialists, not only from within Balfour Beatty, but also customer and development partner London Underground, which is currently in the process of rolling out the solution across its entire network, following an extended and successful evaluation programme.
A crucial aspect of DataMap is its ability to locate data accurately from many disparate recording systems, collected at different times, to produce a single ‘run on run’ view of the whole network which can aid maintainers in assessing developing issues or faults within their area of responsibility.
A better solution
AssetVIEW and DataMap have evolved, from their original functionality of collecting data and providing visualisations of what was and is happening to an asset, into providing the prediction or foresight of what may happen in the future. Balfour Beatty has pioneered the ability to take historic data and apply prediction algorithms to the data which provide an insight into when an asset would otherwise begin to report issues or fail.
This simple but powerful change in approach allows the maintainer to plan where and when to schedule maintenance. A well-planned resource is far more cost effective, and safer, than having one on standby that can be applied in a reactive manner.
Possession planning can become far more effective when specific types of maintenance are required, even combining disciplines in a single possession to minimise disruption.
Network Rail is also currently embracing predictive maintenance systems with the National Gauging Database. Gauging data, clearances and stepping distances for every rail vehicle in the UK are stored in the cloud, accessible via Balfour Beatty’s rail portal (www.bbrailportal.com) on any desktop or mobile device. Data is constantly updated and prediction algorithms indicate when clearances are likely to become problematic, providing engineers with powerful tools to actively manage clearances and reduce inspection frequencies.
Up until now, the primary objective of ACM has been to mitigate or eliminate the risk of an unplanned failure, the consequential disruption to service and the cost of any reactive measures. A worthy aim but, realistically, risk will never be completely eliminated, not without seriously and adversely impacting operations and cost effectiveness. So consideration should be given on how to embrace risk and treat it, not as something to be exterminated, but as a factor to be understood.
Many industries use stochastic methods to provide simulations of likely outcomes – not just banks and high street retailers selling ice creams in summer but real engineering industries such as nuclear power and aerospace. If they can develop models to predict failure or, to put it another way, provide levels of confidence of successful operation to defined limits, then perhaps the rail industry should consider evaluating it too.
If the level of confidence of an asset performing to specification until a given point in the future was available, and there was consideration of the impact to the business if that prediction were wrong, then an informed decision as to whether to continue to operate the asset or take it out of service to effect maintenance could be made. That decision having been taken in a manner that was risk focussed not risk averse, with the potential added benefit of increased asset life and improved safety.
These models are constantly evolving and, to that end, Balfour Beatty and its key strategic partners are embracing that challenge in the field of gauging and overhead line electrification.
Gauging is the practice of modelling vehicle dynamic movements against measured infrastructure to determine the ‘absolute’ value of physical and electrical clearance between the two. Until recently, this practice, as implemented in the UK’s preeminent clearance software ClearRouteTM, has adopted just such absolute techniques in an attempt to eliminate the risk of contact between vehicle and infrastructure. However, it is now possible to utilise sophisticated Monte-Carlo based techniques to provide not only a calculation of the clearance value, but also a measure of the level of confidence the modelling has in that clearance value, a first step to embracing the risk.
At present, this stochastic technique is available as a consultancy service to customers but the plan is to ultimately have it available via Balfour Beatty’s rail portal alongside the other aspects of the National Gauging Database project being delivered to Network Rail.
Asset Condition Monitoring is an evolving, growing and ever changing landscape that has firmly entered the digital age with the advent of cloud technologies, the Internet of Things and advanced software forecasting. It shows great potential to improve safety, extend asset life and reduce costs for the entire rail industry.