From the computer to the rails

Imagine boarding a train that has already covered more miles than your car ever will - without moving a single wheel. In a virtual proving ground, Siemens engineers have already subjected it to comprehensive tests. This lowers costs, facilitates predictive maintenance – and ensures a punctual and pleasant train experience. 

The bathtub curve has a bad reputation. It looks like a long, horseshoe-like “U” and visualizes the failure rate of technical products. Example, the rail industry: In the beginning, when a newly developed tram or light-rail, regional or high-speed train goes into operation, technical hiccups and undetected systematic failures hamper operations or even cause complete breakdowns.  

From Bathtub Curves to Flatlines 

When - for example - the first-generation Highspeed trains launched in the '90s, engineers spent years addressing software bugs. Today, comprehensive digital stress tests prevent these issues from occurring. Frank Hoffmann, Head of Engineering at Siemens Mobility, says:

“Digital twins enable us to spot and fix systematic defects like software bugs at an early stage. Particularly at the end of a product’s life cycle, we can use predictive maintenance to avoid doing unscheduled maintenance work.”

This means one thing for the bathtub curve: It flattens out. 

Overcoming the technology gap: The rail industry is the technology leader 

 “Twenty years ago, people kept saying one thing: When it comes to digitalization, the rail industry lags miles behind, particularly when compared with the automotive industry, a sector that developed better and better driver assistant and entertainment systems decade after decade. Today, we are their equals and have even moved to the forefront in some areas."

Hoffmann says.

Government officials now verify a product's functional capability through today’s life-like simulations. These products include things like the design of air-conditioning systems. Just a few years ago, these systems had to prove themselves during tests conducted in climate chambers. Today, the simulation calculates airflow and temperature distribution and determines whether virtual passengers can keep their cool during the trip. Only one single train of a new platform now has to make the journey into the climate chamber. All other variants based on this platform will be tested and authorized during computer simulations. This significantly accelerates the development process and lowers costs because fewer defects have to be addressed in the real world.

Today, digital twins simulate all components in a train, such as the stress on mechanical parts like bogies or the control behavior of the power electronics. In doing so, developers have also considered a number of details, such as interference currents generated by modern converters that can destabilize the network. And, of course, the software.

“It is 100 percent stress tested.”

Hoffmann says.  

New role assignment: Siemens takes over the maintenance 

One reason that digital twins are so realistic today is the new assignment of roles that are played by the manufacturer, that is, Siemens, and customers, the rail operators. In the past, the customer was responsible for maintenance after a train was delivered. As a result, the operator did not share operational and condition data with Siemens. Frequently, Siemens did not learn of a train’s technical problems until long after the fact or not at all. Today, private train operators in particular have concluded service agreements in which Siemens assumes responsibility for keeping trains in operation – and in return, Siemens gets access to the operational data.  

This data makes all sorts of things possible, as the Rhein-Ruhr-Express, for which Siemens erected its own service center in Dortmund, demonstrates. The trains are packed with sensors that continuously report on the condition of critical operational components. One sensor, for instance, measures the electrical current in the motor that operates the sliding door for the toilet. A deviation of the electrical current from the defined value could indicate a potential defect.

Siemens collects such operational information from all of its modern trains, in a matter of seconds and worldwide. The data flows into Railigent X, the platform for digital train operations at Siemens. Working in combination with digital twins, this data lays the foundation for innovative applications of artificial intelligence. The AI learns from this information and makes better and better forecasts. The AI will identify a potential breakdown of a critical component, and the software will automatically plan for the part to be replaced during the next maintenance window.  

Human beings are simulated, too 

But one uncertainty factor remains even if the technology works to perfection: human beings. Knowing this, many customers obtain a train driver simulator, a life-like operator’s cab in which drivers can train and tackle challenging, pre-programmed scenarios. Siemens’ cab simulators are so immersive, drivers swear they’re real - like training on the latest flightsim setup. Should new operational scenarios be required, Siemens engineers can remotely access the operators’ cabs installed around the world and add new content to them. But this is just an intermediary step. Siemens has developed an automated quality assurance system based on digital twins. This system simulates all operating steps in the driver’s cab – and thus tests the handling of the train from top to bottom.  

Frank Hoffmann views Siemens as a leader in digitalization.

“We started using digital twins and AI much earlier than others. Competitors rely on suppliers. But we retain an extensive amount knowhow about subsystems in our company.”

A clever strategy is responsible for one other factor that sets Siemens off from its competitors. It is a decision that pays off in simulations and digital twins. Oliver Klar, the Head of Power Transmission at Siemens Mobility, says:

“We understand the entire train and deliver higher and higher quality even as trains become increasingly complex.”

No matter whether you are talking about trams, locomotives or high-speed trains.

Automation: coming to engineering, too   

To keep pace with this complexity, more digitalization will be needed, particularly at the beginning of the life cycle, in the development phase. Frank Hoffmann says:

“We already generate software automatically today. We will also significantly increase the level of automation in engineering with artificial intelligence. It is a development that will create new and challenging jobs for engineers.The rail industry, especially Siemens, remains an exciting employer.”

Think about your daily commute – what’s the one train hiccup you’d love us to predict and prevent? 🚆 Was it a stuck door, a sudden AC shutdown, or a service delay? Drop your experience below and let’s explore how digital twins could make every ride smoother!