Predictive rail asset maintenance, considered to be the next big wave in maintenance strategies, reduces costs; improves safety, punctuality and reliability; increases profitability; and minimises service disruptions by predicting the optimal maintenance timing. Advancements in sensors, data analytics and communication technologies have allowed the capture of crucial data and enable monitoring of mechanical and electrical conditions to improve operational efficiency and evaluate performance against selected performance indicators. Predictive rail maintenance, a win-win strategy, is thriving in the digital age.
Figure 1 depicts the four key stages in predictive maintenance.
Figure 1: Four key stages in predictive maintenance
London Underground, UK
Technology deployments: In 2014, Transport for London (TfL) partnered with technology services company telent to install sensors in escalators, elevators, air conditioning systems and subway tunnels as well as monitor PA systems and closed-circuit television (CCTV) cameras.
IT firm CGI has integrated telent's implementation platform onto a network powered by Microsoft's Azure cloud-based intelligent system service. This allows sophisticated predictive modelling, in which real-time data is used to closely monitor temperature, vibration, humidity, fault warnings and system alerts. The data is available in a central location and provides needed information on mobile applications (apps), via a web browser, and through text alerts.
Benefits: TfL plans to boost efficiency by over 30 per cent in three years (by 2017).
Network Rail, UK
Technology deployments: Network Rail’s Offering Rail Better Information Services (ORBIS) programme was launched in 2012 for predictive maintenance. CSC was the lead technology partner for ORBIS and served as the system integrator.
Under this programme, more than 14,000 iDevices were given to frontline teams and these digitised the work order process with apps such as My Work. This allowed the maintenance teams to gain access to a work bank, which provided information such as the condition history of an asset.
In 2014, the Linear Asset Decision Support (LADS) tool was deployed for signalling, switches and crossings, operational property and electrical power. It allows track engineers to see inspection, fault and condition history of each asset from a single source. The users can analyse the root cause of faults from their iPads and desktops and re-prioritise renewal decisions.
In 2016, aerial survey imagery and LiDAR data was made available for all eight routes of Network Rail. The routes together span 16,000 km. An aerial survey was carried out to capture high-resolution and 3D imagery of the entire network.
Benefits: More than 100,000 transactions are processed each week with the My Work app. The solution processed one million work orders, removing 500 boxes of paper and resulting in an estimated 40 per cent reduction in administration requirements, in the first seven months after the rollout in July 2014. By mid-2015, three million work orders had been processed.
The LADS tool helped in decreasing the whole-life cost of maintaining assets.
Going forward: In 2017 and 2018, the Civils Strategic Asset Management Solution is planned to be deployed under ORBIS. It will combine data from many sources to help in the execution of targeted inspection regimes. It will also be used to develop the integrated network model, which will provide a geo-spatial view of the railway and show the logical relationship between assets and the impact of work undertaken.
Technology deployments: Smart predictive maintenance technologies are being deployed at the Three Bridges depot, which was completed in October 2015. MRX has supplied an automated inspection system with laser scanning for the entire fleet comprising 115 Class 700 Desiro City electric multiple-units (EMUs).
Cross London Trains (a joint venture of Siemens Project Ventures, Innisfree and 3i Infrastructure) is supplying the EMUs. Siemens has a separate contract for maintenance.
The depot building has five elevated maintenance tracks (for easy underfloor access). Power at the site is sourced from third rail (750 V DC). One of the tracks has a 25 kV/50 Hz electrification system for testing the dual-voltage train sets. There are 11 sidings providing stabling for 172 cars.
Going forward: Siemens will use its experience at the Three Bridges Depot to develop a facility in Dortmund, where it will maintain double-decker EMUs ordered for the Rhein-Ruhr Express project.
Technology deployments: Spanish train operator Renfe has deployed Siemens Velaro E high-speed trains. Nertus, a joint venture between Renfe and Siemens, undertakes maintenance on the 26 eight-car units using advanced data analysis. Velaro trains operate in temperatures from –20°C to +50°C at speeds of up to 350 km/hr. A train developing abnormal patterns is dispatched for an inspection service to prevent failure on the track.
Siemens undertakes maintenance at flexible intervals under a performance-based-maintenance contract.
Benefits: Predictive maintenance has resulted in impressive reliability achievements. Only one of 2,300 journeys were delayed, and these only by five-minute delays. Maintenance-related delays longer than 10 minutes only occur every 1.5 million km, which translates into less than 0.0005 per cent of all journeys. Renfe has achieved a punctuality rate of 99.98 per cent.
Technology deployments: French national train operator SNCF has deployed IBM Watson’s Internet of Things (IoT) deep learning analytics platform and the SigFox IoT network across the rail network, which spans 30,000 km, covers 3,000 stations, and includes 15,000 trains. The sensors and cloud enables SNCF to run distributed calculations and re-inject the results into train and rail maintenance processes.
A few key applications are discussed below:
Company strategy: The deployment of technologies is part of the company’s 2020 Vision, to become an industrial champion striving for operational excellence and optimum efficiency by using the IoT.
SNCF’s deployment of digital technologies is guided by three core principles – cyber security by design, platform as a service deployment model, and leveraging Big Data for decision support.
Benefits: The new technologies have allowed SNCF to improve operational performance, customer experience and safety. SNCF has estimated that the cost of maintenance of tracks and trains could be reduced by a factor of 10 with these technologies.
Remote monitoring has reduced the maintenance time at train depots and improved the turnaround time of the trains.
Technology deployments: In early 2014, a dynamic maintenance management system using the SAP(R) predictive maintenance solution was deployed to analyse sensor data and monitor equipment behaviour remotely and undertake corrective actions in service and maintenance systems. It allows Trenitalia to link data from equipment such as motors, batteries and brakes with lifecycle models, usage wear and other performance indicators.
Trenitalia has a fleet of around 30,000 locomotives, electric and light trains as well as coaches and freight cars. It operates more than 8,000 trains per day.
Company strategy: Trenitalia aims to provide better services to customers and achieve substantial reduction in industrial costs. It plans to invest in high-tech solutions that enable more efficient, sustainable and fast journeys.
Benefits: Trenitalia expects to reduce maintenance costs by 8 to 10 per cent.
Key recent deployments
Rail operating companies in Europe are mostly government-owned and lack the resources to invest in a comprehensive predictive maintenance strategy. The solutions are mostly deployed in new systems/sub-systems but not to retrofit existing rolling stock.
In addition, rail operators are not able to justify the cost of collecting huge amounts of data, which is a necessity for an effective predictive maintenance solution. As a result, predictive maintenance solutions are only being implemented in a few cases where the leadership has a vision for the future and the capability to implement the necessary changes within the organisation.
Further, climate conditions vary from country to country; therefore, rail network and rolling stock are tailored to specific environments. This implies that an across-the-board predictive maintenance solution cannot be deployed on all networks.
Until the predictive maintenance technology is sufficiently advanced, adaptable and affordable to be implemented on a large scale, it is unlikely that many rail operators will invest in it even though the benefits from the solutions can boost profitability and revenues. More pilot programmes and public-private partnerships can pave the way for the deployment of wider-scale predictive maintenance programmes.