Experiences with machineering/iPhysics?

[Faser Active]

I’m currently evaluating machineering and their simulation platform iPhysics and wanted to hear some real-world experiences from people who have actually worked with it in industrial environments. I saw the programme at a trade fair and wanted to find out more about it.
I’m especially interested in how stable and performant iPhysics is during day-to-day use, what kinds of applications you are using it for (virtual commissioning, robotics, material flow simulation, PLC testing, digital twins, etc.), and how well it integrates with existing PLC, CAD, and robot controller ecosystems. I’d also love to know how steep the learning curve was for your teams and whether there were any limitations or pain points that became apparent during larger projects.
If anyone has compared it directly with some alternatives, I’d be very interested in hearing your perspective on where iPhysics stands out and where it falls short.
Would appreciate any honest feedback

 

[Juls_]

I actually dealt with digital twins in my last job. It was in the food industry, and the company had been doing it for a while. The pros are obvious—you can catch potential process glitches before they hit the actual machine. Definitely leads to fewer errors, less downtime, and so on.

As far as I know, iPhysics was easy to integrate into existing systems. It took a few workshops to get started, but our team was able to run with it on their own after that.

What kind of project are you looking to use it for?

 

[Faser Active]

How many people were on the team, and how long did the training take, roughly?
One of the main reasons I am looking into iPhysics is that one of our customers would like to have a digital twin of a machine that we are currently planning and building as a machine manufacturer. The idea is to create a virtual representation of the machine before it is physically assembled, allowing him to simulate its behavior, validate the design, and test different operating scenarios in advance.
The customer really wants this, but we’ve never created it ourselves before.

 

[Juls_]

That’s the perfect start into digital twins! Honestly, it’s great that your customer is asking for this. Building the machine virtually before you do the real assembly is a total game-changer for both sides.

To give you a rough idea about our setup: we just started with a small team of 2 or 3 guys, one doing the CAD/kinematics stuff and one for the PLC programming. The basic training and getting our first model alive took maybe 2 to 3 weeks.

Doing this before the physical build has massive benefits: You connect the virtual machine to your PLC code and find bugs or collisions on the screen. Much cheaper than fixing broken metal on the shop floor. You can simulate heavy stress tests or weird scenarios easily before buying any hardware. And the customer can already see and approve how the machine works.

The learning curve is totally okay because you can just reuse your CAD data. What PLC system are your guys using?

 

[Faser Active]

I agree that building and testing the machine virtually before the physical assembly can bring huge advantages for both the manufacturer and the customer. Especially the possibility to connect the virtual model with the PLC program and identify software errors, collisions, or sequence problems early can save a lot of time and costs during the real commissioning phase. That´s exactly why we are interested in the iPhysics Software. If anyone else has experience with similar programmes, I’ll take those on board too. That was just the first piece of software I came across in this area and the one I’ve already done a bit of research on. I’ve also read that it’s relatively easy to use, so that’s a good start.
Starting with a small team is a good approach because the first step is often more about gaining experience and developing the right processes than having a large setup from the beginning. Reusing existing CAD data is definitely a big advantage, as it reduces the effort needed to create the virtual model. In addition, involving the customer early by showing them the machine behavior virtually can improve communication and help avoid misunderstandings before the machine is built. Especially when you’re dealing with customers who have no idea what it’s all about. I heard that happens too
The possibility to test different scenarios or stress situations without risking damage to real hardware is another important benefit. This makes it much easier to optimize the machine and increase the reliability of the final system.
Regarding the PLC system, we are currently working mainly with B&R, but we want to switch to another provider.

 

[Juls_]

I think you've summed up the main benefits really well. In my opinion, the biggest advantage of virtual commissioning is that you can test the PLC software and the machine sequences before the actual machine is even built. That way, a lot of issues can be found and fixed much earlier, which makes the real commissioning process much smoother and less time-consuming.

Once the digital twin is connected to live operating data, it can also be used for process optimization or troubleshooting. That's where I think it really starts to add value.

I don't have any experience with other software yet. With iPhysics, they offered workshops at the beginning to get everyone familiar with the software, so you can start working with it on your own pretty quickly.

 

[Faser Active]

I also think one of the less obviuos advantages is that different teams (mechanical, electrical and software) can work in parallel instead of waiting for the hardware to be finished.
I'm still curious how well that translates into day-to-day projects, though. Does the time spent building and maintaining the digital twin usually pay for itself, especially on smaller or less complex machines? That's something I'd be interested to hear from people who use that kind of software regularly.

 

[Juls_]

For smaller or less complex machines, I don’t think it’s worth modeling every little detail digitally. The time and effort it takes to build and maintain the digital twin still needs to be justified by the potential benefits.

It gets more interesting when you can use the digital twin to test different scenarios, variants, or failure cases before doing it in the real world. That can quickly pay off by saving time on prototypes, modifications, and troubleshooting later on.

And for larger, more complex projects, I’d expect the benefits to be even bigger.

 

[Faser Active]

I'm not raelly familiar with how these projects are handled in practice, so I'm genuinely curious about where the tipping point is.People often say that digital twins are only worth it for larger systems, but what actually counts as large or complex in this context? For me, it always makes sense when you can expect problems to come up. For me, it’s not a question of how amny problems there might be, but whether problems might arise. Of course, with larger and more complex machines, this is more likely to be the case, but that doesn’t mean problems can’t occur with smaller or not so complex machines.
I also wonder how much work is involved in keeping the model up to date once the real machine changes. If the design is revised several times during development, does the digital twin stay useful, or does it become another thing that needs constant maintenance? Is there a point at which you’d say it’s no longer worth it? From my point of view, not really

 

[Juls_]

I see it pretty much the same way. I wouldn’t judge whether a digital twin is worth it purely based on the size of the machine. What matters more is the risk and cost involved if something goes wrong or changes late in the project.

Even with smaller machines, virtual commissioning can make a lot of sense if you have complex sequences, lots of interfaces, or challenging PLC logic. A smaller system can actually require more testing than a much larger but highly standardized machine.

When it comes to keeping the model up to date, I think a modular approach is really important. If the digital twin is built around reusable modules and functions, a change doesn’t automatically mean rebuilding the whole thing. Ideally, you just update the part that was actually changed.

The problem starts when the digital twin is modeled in way too much detail and every small mechanical change creates a lot of extra maintenance work. That’s why I think you should define from the beginning what actually needs to be represented for the virtual commissioning. The model doesn’t have to look photorealistic, it just needs to reproduce the relevant realworld behavior well enough for what you’re trying to test.

That’s also why I wouldn’t set some fixed point where a digital twin suddenly stops being worth it. The more expensive it is to find and fix problems late in the project, the more valuable the digital twin becomes, regardless of whether the machine is big or small.

 

[Juls_]

.

 

[Faser Active]

I think that is a fair way of looking at it. In the end, there probably isn’t a simple rule based on machine size, because the valeu of a digital twin depends much more on complexity, interfaces, project risk and how costly late changes are as it seems. The level of detail has to be chosen carefully. A model that is unnecessarily complex can quickly become a maintenance project in itself, which defeats part of the purpose. You’ll first have to work out where the line is between what’s useful and what’s too much.
The modular approach is probably one of the most practical ways to make the whole concept sustainable over several projects. If the relevant functions and components can be reused, the effort for both initial modeling and later updates becomes much more manageable. For virtaul commissioning the important thing is not how realistic the model looks, but whether it behaves realistically enough to test the relevant processes and control logic.
So I think we are actually quite close in our assessment. There is no universal point at which a digital twin makes sense or doesn’t make sense. It is ultimately a question of whether the additional effort is justified by the risks and potential savings during commissioning and later changes.
Thanks for the nice and informative conversation.