One thing I adore about my job is being able to spend time in factories. There’s something really satisfying about the hum of machines as they work in unison to create products. To me, a factory is like a fine mechanical watch.
When all the gears are working in unison, you’re good to go. But if even a slight change is made to just one element, everything comes to a halt. Then it’s time to consult a skilled watchmaker. This has parallels to how a factory works. I’ll show you an example:
At a Siemens factory in Erlangen, control mechanisms from the SINUMERIK, SINAMICS and SIMOTION product families are all constructed on the same line. Orders for a specific product are collected and then the products are made in batches.
For product changes, potentially 36 parameters have to be adjusted on the line. And like with a mechanical watch, changing just one element requires countless other modifications. These 36 parameters beget an exponential number of configurations – somewhere in the neighborhood of one billion!
Traditionally, operators make adjustments manually, consulting Excel charts containing the settings for each product. They also rely on their experience and gut feeling. But as you can imagine, such changes are time-consuming and carry risks.
At one point, my colleagues and I were called in to see how digitalization could help optimize the plant. We needed to ensure the highest level of efficiency with each product change. For assignments like these, we follow four steps:
1. Digital twin with Plant Simulation
The first thing we need to do is gain a complete understanding of the production processes and pinpoint the challenges. To do this, we develop simulation models of the facility and construct a digital twin of production using Plant Simulation:
see here the real plant…
and the virtual one.
Now we’re ready to run simulations. We can test and evaluate different production scenarios virtually. This lets us discover bottlenecks in production, among others.
3. Design Exploration with HEEDS
This step is all about finding the optimal solution while taking all parameters into consideration. Here, the Digital Twin is connected to analysis algorithms that leverage AI and data analytics. HEEDS – the software tool – generates and tests all different combinations possible on a production line.
Now we’re getting to the really exciting part! We evaluate the parameter sets that HEEDS comes up with in the real production environment. The really cool thing is that we already know exactly what we need to do because everything has been tested virtually. For the factory in Erlangen, our recommendations brought:
- up to 17 percent reduction in material requirements,
- a 42 percent reduction in the number of containers on the line, and
- a 20 minutes reduction of production time every day.
Quite simply, the plant can now get more done with less.
My colleagues and I not only provided the models, we took care of the consulting and implementation too. This is what all the steps – from simulation to implementation – look like:
Aside from optimizing Siemens’ own digital factories, we work with clients from just about every sector, from food and beverage to automotive and to electronics. Each factory might be different, but for me, there’s a constant: the feeling of awe as machines go about their work, much like a mechanical watch.