How do you get a digital twin running in the cloud with ThingWorx? How do you set up an IoT-connected digital twin? And why would you want to do that? Let’s take a few minutes and talk about it. 

Companies might be capturing sensor data off of a product in the field and applying AI/machine learning to it to predict what’s going to happen in the future. But some companies want more insight into what’s going on with that product in the field. One way to do that is to set up a simulation where you feed it sensor data, and then you take other measurements off of that simulation. This simulation is often called a digital twin. The idea here is that you gain more insight from those other measurements.

So how do you do that? I want to answer that question by talking about PTC’s products and how they work with ThingWorx. Let’s walk through an example of how this might work. You have an autonomous tractor out in the field that you are streaming data off of that has a lot of different sensors. That data is going up into the cloud to ThingWorx. ThingWorx is collecting this data over time. 

You can mimic the behavior of the tractor with Integrity Modeler. Integrity Modeler can build these 1D system simulations that are almost like diagrams with the inputs and outputs. They’re all equation or formula-based. You can build those models and run those models next to ThingWorx using a headless application. There’s no UI. Basically, it’s an application that’s running things in the background, but it’s getting sensory data from ThingWorx and then feeding these extra measurements back into ThingWorx. So you have a richer set of information where you can gain insight.

The other thing that a lot of people probably are not aware that there’s a similar type of thing for Creo. It’s called Creo Product Insight. For example, you can set up a kinematic simulation. It’s a little simpler. It can’t really be a stress type of simulation because of the time required to run those simulations. But what it can be is this calculation of lots of other things that might be: 

• Where’s the center of gravity? 
• What’s the surface area above water? 

All types of things that are basically modeled-based driven calculations. You can feed this headless Creo Product Insight application that’s running alongside ThingWorx in the cloud. You can feed it sensory data and it drives extra information and data back into ThingWorx.

Let’s be clear. There are very few other companies other than PTC that are making these types of investments to extend the digital twin. This kind of connected digital twin that’s mimicking the behavior of something in the field. This gets into some really interesting applications. You can look at is the tractor that you’re working in — is it about to tip? Is it about to be submerged below the air intake for the engine or something like that?

You can use this to drive other types of warnings or behaviors, or even let AI machine learning loose on it. Let it try to correlate those measurements with other behaviors, or things that affect predictive maintenance and those types of strategies that you want to get in place for your company.

So this is something where, there are some complex aspects of your design, that as you’re going through the engineering aspect of your product development process, that say, okay, these are some key measures for behaviors. This is how we figure out if we’re on target or not. It’s not out of reach to say, hey, let’s measure that over time. Let’s use that in a headless Creo Product’s Insight instance of application in a model. Let’s see if this actually does make a difference. Let’s close the loop on our hypothesis of how this thing is going to be used and behave, and see if that actually is the case.

It’s a really interesting application. It’s going to be useful for some companies. If it does seem like you’ve gone through that, there are some key characteristics that you think will make an impact and you want to see if it does or not, this is a good place to experiment and check that out.