Centralised diagnostics aids transition to predictive maintenance
This article is written by Kevin Smith and was originally published on International Railway Journal on June 12, 2018.
Remote monitoring equipment developed by Railnova is providing a centralised system for fleet diagnostics and is now in use on more than 1000 locomotives and trains across Europe. It is also supporting the shift to predictive train maintenance, as Kevin Smith explains.
As train operators and fleet owners consider how they might transition from traditional time-based to condition-based maintenance practices, they are faced with an array of possible diagnostics solutions.
Individual sensors are now available in the marketplace to measure anything from fuel consumption to component temperature, engine performance and battery condition. The Internet of Things (IoT) offers great promise to railway asset owners and operators. Yet harnessing this information and transforming it into relevant actions remains a major challenge.
Before the era of widespread fleet connectivity, downloading this data was a cumbersome task: fleet managers typically went to the individual locomotive or train and inserted a USB stick into a diagnostics port. At best, this data was then sent to a central location where, in the majority of cases, it is only stored per component and rarely analysed in a timely manner to identify anomalies. Diagnostics PCs are usually required for each manufacturer and each sub-component to access and read this specific data, adding to the complexity of the task. The expertise required is also usually restricted to a few individuals within the company, limiting analysis and understanding of how this information might be used. In addition, the nature of the process means that the information is not delivered in real-time, preventing drivers from troubleshooting issues while they are on the move or preventing fleet engineers from anticipating failures from patterns visible in the data.
Spotting an opportunity in the market to deliver a universal telematics solution that can overcome the complexity of this data-gathering process, Mr Christian Sprauer founded Railnova in 2010. Sprauer had previously worked for Alstom and rolling stock leasing company Alpha Trains, and alongside his former trainee, Mr Charles Henri-Mousset, who had built a working telematics prototype in six weeks, the pair began to develop a viable product.
The company quickly secured its first customer, rail freight operator Europorte, in 2011, and its solution now covers more than 1000 rail vehicles. Among the start-up’s leading clients are Alstom, Lineas, German Rail (DB), French National Railways (SNCF), SBB Cargo, Alpha Trains, VFLI, ETF and Netherlands Railways (NS).
At the heart of Railnova’s solution is Railster, an onboard railway-certified and powerful edge computer with FPGA technology. Railster hardware is installed onboard a vehicle and connects to the available data sources such as MVB, CAN, Profibus networks, central control unit, door controllers, HVAC controllers, engine controller or onboard sensors, retrieving data and measuring various component parameters in real-time. On passenger trains and LRVs, further Railnova modules connect to sub-components located in the vehicles to gather information, which is transmitted wirelessly to Railster.
Railnova’s country director for Germany, Austria and Switzerland, Ms Babette Müller-Reichenwallner, explains that Railster is able to continuously and effectively filter the data collected from equipment used on any type of locomotive or train.
“The solution can be installed during any maintenance interval as it only takes half a day for two technicians to install what is an ‘Ikea-style’ kit,” Müller-Reichenwallner says. “During deployment the first installation is done by us together with the customer installer, and then it is carried out by the operator. It is also a very flexible product; our customers are encouraged to start by monitoring the components that are most prone to disrupt services and add more later.”
In addition to the Railster, Railnova provides a complete end-to-end solution with its data management platform, Railfleet, able to act as the back-end for the data gathering process for an entire fleet of locomotives, trains and LRVs. Through Railfleet, the user is able to select the parameters to measure, with the backhaul communicating “over-the-air” with the Railster installed on the rolling stock. It is possible to relay this data to a Big Data analytics platform such as IBM’s Watson IoT or Predix. Railfleet also acts as the main user-application, where fleet engineers can support distressed drivers in real-time, manage and analyse any fault codes, and turn these into work orders. This data can subsequently be fed into an enterprise resource planner (ERP) or maintenance management system (MMS) via Railfleet’s standard API.
Müller-Reichenwallner adds that the system helps to quickly identify any anomalies in the data and it is possible to relay this information in real-time to drivers, the control centre and teams on the ground via Railfleet, which is web-based and therefore accessible through a desktop PC, tablet or mobile phone. The information can similarly be shared in real-time with clients and maintenance partners.
Rule engine
Critical to Railnova’s application is a rule engine, a software component that applies logic to multiple data sources to detect events, which the client feels are particularly relevant, and can also prevent spam from others which are less significant. Railnova says the rule engine helps fleet engineers to build asset intelligence independently from the OEM while using edge computing reduces the pressure and cost of running servers to download vast amounts of data.
Rather than simply selling boxes, Railnova’s business model is based on offering its software as a service. Railnova is responsible for server hosting, back-ups, software maintenance, migrations and security. Collected data is stored securely at Railnova and made available in the cloud. Müller-Reichenwallner says all data remains the property of the client with Railnova only managing and accessing this data for the purpose of service delivery, with the company obliged to return it at the end of the project or on demand.
As the client base has expanded, the technology has matured, progressively helping to measure more parameters and aiding fleet maintenance practices. For example, the inaugural project for Europorte involved the installation of GPS, fuel consumption monitors, and an engine hours and fault code monitoring system on a mixed fleet of 50 units, which enabled the operator to dramatically improve its fuel consumption. For other customers, the supplier is harnessing the data retrieved from sensors measuring oil pressure, coolant temperature, battery voltage, kilometres, speed, idle time, and pantograph position. It has also developed a system for real-time monitoring of ERTMS, which is in use on Bombardier Traxx locomotives used by NS.
Key components
Müller-Reichenwallner says one of the key components where monitoring can make a particular difference is the locomotive battery.
“We always connect to the battery,” Müller-Reichenwallner says. “It is a component which only costs around €1000 but without it, you can’t run the train. If customers invest in the system, we estimate that they can receive a complete return on investment in one winter because of the improved reliability of the battery alone.”
Through enhanced component and system condition monitoring, Railster and Railfleet are providing the foundation for a transition to predictive maintenance. Yet Müller-Reichenwallner says that fleet owners must be mindful of the practicalities of maintenance and its specific response to the alerts from the system, particularly for problems which will not hinder operations in the short term. “It is all about finding the golden path,” she says. “Otherwise they run the risk of the vehicle being in the workshop everyday.”
Railnova says it can overcome this challenge by offering prospective customers the opportunity to develop a proof of concept. Indeed, Müller-Reichenwallner says its engineers are able to meet a client’s needs more effectively than by responding to a request for qualification. For example, since securing a major contract with DB in 2015, Railnova has installed its technology on 220 Bombardier Traxx locomotives and 16 Vossloh G1000 diesel locomotives. However, the project started as a prototype in 2012, which was expanded to a proof of concept on 10 locomotives for a full year of operation in 2013.
“We are always looking to improve our offer based on customer feedback,” Müller-Reichenwallner says. “Every four to eight weeks there is a new release of the system, where we are improving the requirements and putting it into production. However, the level of service always remains the same.”
This approach is also helping the supplier to overcome internal challenges. Müller-Reichenwallner reveals that as a start-up working in the railway industry, which is renowned for its conservatism, Railnova is facing a cultural as much as a technical challenge from prospective customers in the market. The company has made great strides in the past eight years. However, there are still some important barriers to overcome if the industrial internet of things is to take off in the rail sector.
“The biggest challenge we face is the adoption of the system internally,” Müller-Reichenwallner says. “Is the organisation really ready to do this? Does it have the right people or enough people to commit to it? Many will require a major internal transformation to profit as much as possible from these types of solutions. We are ready and happy to work with them on this.”
Interested to learn more, or want to receive a quote? Get in touch, or take a look at our website.