TurboTime Podcast: Part 2: Applications/AI for Digital Twins

In part two of this episode on digital twins, Myth Busters Klaus Brun and Rainer Kurz expand on the applications of a digital twin, its advantages and limitations, and the integration of AI in the platform.

Rainer Kurz highlighted a best-use scenario for digital twin technology: “If you have large fleets of reasonably similar machines, it’s much easier to set up a system of digital twins for condition-based maintenance where you get constant data flow and are able to advise the users on maintenance decisions. It’s actually a very powerful tool.”

Brun added his thoughts on modern digital twin integration: “The combination of a digital twin with other technologies such as fleet monitoring, forecasting, and condition-based monitoring provides an opportunity to improve the operation of the machine.”

Brun said he’s seen promising work being done with AI. “I recently read a doctoral dissertation from a university in Europe, which was interesting because AI was able to parse the data and come up with some intelligent forecasting of failures and at least, theoretically, improved the availability of the machine,” he said. “Learning algorithms have been around for about 30 years or so, but they’ve really only gotten good in the last five years. Now we’re starting to see AI make inroads into more complex things like compressors, gas turbines, steam turbines, plants, etc. I have no doubt that we’ll see more of this and better AI, but at this point in time it doesn’t replace any operator and it’s in its infancy.”

Kurz continued the discussion on AI applications: “One of the big promises of AI is guiding an operator through the system he’s operating. For example, helping the right warnings to show up at the correct time and giving him the right information. There is not AI that simulates complex systems, it simply augments the capability of the human to help make decisions.”

Click here to listen to part one.