Siemens Worldwide

Pictures of the Future



Mr. Sebastian Webel
Mr. Sebastian Webel


Mobile: +49 172-7169762

Werner-von-Siemens-Straße 1
80333 Munich

Pictures of the Future
The Magazine for Research and Innovation

Simulation and Virtual Reality

How to Step inside a Gas Turbine

CT researcher Jan-Philipp Fahlbusch with a virtual reality headset. Siemens experts use VR to scrutinize the digital twin of a gas turbine down to the smallest detail in their search for functional anomalies.

Researchers at Siemens are developing virtual copies of gas turbines that promise to revolutionize predictive maintenance and repairs. Known as digital twins, these realistic 3D models, which are visualized with virtual reality goggles, are based on up-to-the-minute sensor data and the use of artificial intelligence.

Siemens’ latest gas turbines contain more than 500 sensors, which continuously register pressure conditions, temperatures, component stress, and much more. The resulting data is evaluated by software, generating information that helps turbine operators easily track parameters that are crucial for operation.

But that isn’t the sensor data’s only benefit. Engineers can also use it to create a digital twin – a copy in the virtual world – of a turbine in order to remotely monitor the system and solve maintenance problems early on.

And this is exactly what experts at Siemens Corporate Technology (CT) have successfully done in 2017. In their ongoing search for new data sets for improving their innovative algorithms, the experts found what they were looking for at Siemens’ Power Generation Services Division, which supplied them with the sensor data and operational records of a state-of-the-art Siemens gas turbine.

In a virtual reality representation of a gas turbine, complex sensor data is translated into colors to make the meaning of information – in this case temperature differences – easy to understand.

The Future of Remote Maintenance

That was a stroke of luck. “Many people are unwilling to share their data,” says Ulli Waltinger, who heads the CT virtual reality research team. “But this concern is unwarranted, because we ensure that all data is kept confidential and anonymized before we begin to use it for our research.” Such data sets are needed, for example, to improve algorithms for preventive maintenance, image recognition, and functional optimization. “We need them in order to develop new applications,” adds team member, Felix Buggenthin, an expert in machine intelligence.

CT researchers have demonstrated what the remote maintenance of complex machines might look like in the future.

Using sensor data, a 3D model of a gas turbine (supplied by the Power and Gas Division), and their in-depth knowledge of control centers and maintenance work, CT researchers have now been able to demonstrate what the remote maintenance of complex machines might look like in the future.

Shaping the future of remote maintenance. Ulli Waltinger (center) and colleagues Felix Buggenthin (left) and Jan-Philipp Fahlbusch in their Munich lab.

The Color of Wear and Tear

After donning virtual reality goggles, researchers step into a virtual hall that houses a turbine’s virtual twin. There, sensor data that is supplied by a computing cloud depicts the turbine’s operating status, including information about its combustion temperature and rotation speed, collectively forming a multidimensional information complex.  “We began by doing various trials with the data. The challenge was to depict all of the sensor values in a way that was easy to understand,” explains Jan-Philipp Fahlbusch, who is responsible for the development of the virtual reality environment.

In this effort, the researchers benefited from the fact that digital spaces can be manipulated in almost any way. Due to the spaces’ virtual nature, it’s possible to use a kind of “X-ray vision” to penetrate machine interiors, for example, or to change the colors of components in order to depict changing temperatures or speeds.

This, in turn, enables Siemens to simulate a turbine’s current operation and use artificial intelligence programs to depict other data that is not easily accessible. Examples include the use of colors to show the expected wear and tear of a component so that prompt maintenance measures can prevent unnecessary downtimes.

If a simulation that has been fed with real-time turbine data is placed into the cloud, maintenance engineers can enter the digital space from any place on earth in order to jointly solve problematic situations, thus opening completely new opportunities for Siemens’ portfolio.

Susanne Gold