WITH railways and infrastructure managers spending millions on high-output track maintenance machines, it is imperative that these assets are used to their full potential and are not left idle beside the track.

This problem is particularly acute in large countries such as India, the United States, Brazil and Australia, where infrastructure managers are overseeing vast networks and the location and work of their maintenance equipment is not always clear.

ptc datascientistFor years, Plasser & Theurer track maintenance machines have accumulated data on their activities and performance through the PlasserDatamatic platform. However, collecting the data is one thing, analysing it in order to inform the decision-making process is a completely different and far more complex undertaking.

Big data analytics is one of the key building blocks of digitalisation and the foundation of desirable processes such as predictive and condition-based maintenance, which can potentially bring added value to customers and users of the machines. However, Plasser & Theurer did not possess the skills within its existing organisation to harness this information effectively. As a result, the company created a new subsidiary, separate from its existing operations, and populated with the talent able to develop the required solutions.

P&T Connected is located in the north of Upper Austria Province, at the University of Applied Sciences Upper Austria’s Campus Hagenberg, the so-called “Austrian Silicon Valley,” or what Mr Bernhard Maier, chief operations officer of P&T Connected, prefers to call a “Silicon Village.” The company’s headquarters sit alongside 70 similar small IT and software development start-up companies, which are closely-aligned with the university and its 1500 students.

While serving a variety of sectors, engagement between these companies is encouraged. “We have a communal kitchen and we often have events where we meet and share ideas,” Maier says. “It is about promoting a sense of community thinking. We might be working in completely different areas but many of the problems we have to solve are very similar. We can ask questions of what approach someone else might have taken, and with thousands of different solutions possible, it is likely that someone within the community will have the right answer.”

P&T Connected has come a long way since it was founded in January 2017. Maier describes the company as starting out in an empty office without even a table, like a true start-up. Just five people were employed initially, which has now grown to 13, although the company retains its start-up philosophy in its approach to work.

“We have no staff from the parent company and all employees are software developers and data scientists, new in the business of track maintenance,” Maier says. “They are young talented people, straight out of university, with no influence from other companies. It is also a multicultural team from five different countries and the common language here is English, which is unusual for an Austrian company.”

The team is tasked with developing cloud-based software for track maintenance machines, which are offered under the PlasserSmartMaintenance umbrella of products. These comprise apps based on networked cloud solutions and web services, which aim to provide straightforward access to relevant data on specific machines, fleets and infrastructure.

This includes updating PlasserDatamatic, which, using GPS data, engine data, filling levels and working parameters, is now able to provide users with detailed machine information, reports and analysis via a browser-based user interface. Specific features of the enhanced product include:

  • Last events: a machine’s log entries are accessible and retraceable for a week
  • Integrated user help: all data on the machine is available via a single access point with operating instructions stored within PlasserDatamatic
  • Personalised dashboards: each user is able to select specific machine parameters that they wish to monitor in a customisable display, and
  • Servicing: centralised access to reports of machine utilisation and MachineMaintenanceGuide (MMG) reports, including data such as photos, check lists, notes and audio files.

Prototype

A prototype of the system is already undergoing trials with BNSF in the United States and JR East and its subcontractors in Japan. Installation is also set to begin with a Brazilian railway this month. And while the solution is purely desktop-based at present, Maier says Plasser & Theurer plans to show tablet and smartphone versions of MMG at InnoTrans.

“When entering the web portal, all of the information available for that specific machine is visible,” Maier says. “The interface is fully responsive and easy to use, even for those of us who are not addicted to computers. There is a basic ‘machine view,’ which includes one page of information and an overview of the machine’s current performance, and a ‘technical view’ where you can go into more detail.”

While all Plasser & Theurer machines are customised to the specific needs of the individual customer, Maier says PlasserDatamatics is suitable for anything from a large tamping machine to a small stabiliser. Inevitably the amount of data generated varies by machine, but he says that 50MB per hour of data is typically extracted from the largest machine at the top end of the application.

Appetite

Feedback from these early users has been good according to Maier. And while he admits there is a reluctance among some infrastructure managers to share their data, in general there is an appetite in the market to make better use of this information and make the transition to increasingly predictive and condition-based maintenance of track maintenance equipment.

Developing these specific applications is the task of the data scientists employed at P&T Connected. Maier says that they are currently engaged in various offline projects to harness the data and develop specific tools that facilitate predictive and condition-based maintenance practices. Many of these will utilise machine learning techniques which, as they evolve, will produce an increasingly automated and reliable system and will form the basis of a prototype. “Once it is stable enough and showing around 95% predictability we can bring it to the customer for a trial as a new software solution,” Maier says.

Among the new products currently under development is a condition-based monitoring system for tamping units, which Maier says will also be on show at InnoTrans.

He adds that various other projects are underway. And in areas where the data scientists are short of specific railway expertise, they are drawing on the experience of Plasser & Theurer’s electrical and mechanical engineers, who Maier says are readily available and willing to share more than 20 years of experience in this field.

While the experienced engineers are providing vital railway expertise to the P&T Connected employees, equally the new software developers are bringing new ways of thinking and approaching problems with traditional rail industry systems.

Indeed, Maier is convinced that the new ways of thinking and working in place at P&T Connected will become more and more integrated into the parent company and act as a means of preventing the company from becoming inward looking, or as German speakers say, “wearing internal glasses.” He believes this approach is essential for Plasser & Theurer to move with the times, and with rail infrastructure under greater pressure, deliver what its customers demand: more efficient and reliable track maintenance equipment.

“For me it is important to blow up the traditional knowledge pyramid,” Maier says.

“If a machine is standing still it is costing everyone a lot of money. We are now working on things that will help our customers to benefit from this new data-driven approach to track maintenance and improve the productivity and reliability of their track maintenance equipment.”