Simulations of production facilities help the operators determine the ideal operational workflows. As an integrated part of a digital twin, these simulations can draw on current data flows and swiftly find clear options to help the system operators reach a decision when problems arise.
Something’s amiss in the Fischertechnik mini-factory at the Siemens Digital Machines Lab in Munich. It’s making rattling, humming, beeping, buzzing, and droning noises. Blue, red, and white pieces are being transported from a high rack along conveyor belts, where robot arms sort them by color. Blue items are put to one side in a buffer, while red and white continue to a multi-process station, where they are heated, drilled and milled. But while the building pieces are jiggling through the factory like this, Siemens simulation expert Jan Christoph Wehrstedt, who is standing alongside, notices a small red light flashing on his tablet. What’s up? There’s a potential bottleneck just ahead of the multi-process station, which could bring the production flow to a standstill.
But why? A simulation running in parallel, which constantly compares the actual situation in the plant against the theoretical situation, has identified the problem: The red pieces need longer than planned to heat up. At the same time, the simulation has calculated a way of structuring the production process so everything will run smoothly again. A window appears on the tablet offering this option. Wehrstedt clicks on it, the little red light goes out, and the gray, red, and black factory rattles on.
Assistance system suitable for non-experts
Real automated production facilities also experience problems similar to those in the mini-factory. Not only do they have to operate around the clock and meet quality, time, and cost requirements, but more and more often flexible, small production runs are also required. That means the workflows in a facility like this have to be regularly adapted, and must remain optimized to avoid bottlenecks like those in the mini-factory, for instance. That’s a complex problem for the operators, since changing the workflows may have an impact only some time later, at a completely unexpected location. This is where simulation models integrated into the production system can help.
A team from the Simulation & Digital Twin Working Group at the Siemens Technology Research Unit therefore developed an intuitive assistance system that uses a model of this kind, giving shift leaders, maintenance engineers, and technicians the opportunity to respond to production problems at short notice. “The app is intuitive, and offers clear options regarding ways forward,” says simulation expert Annelie Sohr, who works together with Wehrstedt. “The complex simulation tool, which was previously thought of as being only for experts, runs in the background and shouldn’t be anything for the users to worry about.”
Fully automatic in the long term
These real-time simulations are fed with all kinds of sensor and operating data, enabling rapid error analyses and “what-if” scenarios. In other words, the threatened bottleneck is identified before the blocks start stacking up. “In our mini-factory, for example, we received the suggestion to change the sequence in which the pieces are processed,” observes Sohr. “That means the plant manager receives the suggested option – to make adjustments to production – in super-quick time.”
Of course, it’s also possible to imagine expanding the data sources still further. That would enable the simulation to include the order situation in the coming weeks, inventory levels, or forthcoming maintenance work in its simulation. A precondition, however, as with all data analyses, is that all the relevant data must be available in clearly defined formats. “In the longer term it’s possible that most of these decisions will be taken automatically,” says Wehrstedt. “Operators would then simply formulate general targets, relating to quality assurance or resource-efficient production, for example.”
Already in use
Parts of this assistance strategy are already in place, in an infrastructure environment, for example: The Siemens sewer network control system app SIWA Sewer thus uses simulations to help optimize the operation of drainage systems. In a pilot project in a sewer network in North Rhine-Westphalia, it utilizes a simulation model that draws on resources like weather forecasts. That means SIWA Sewer can ensure the network continues to function smoothly even under heavy rainfall.
Source : SIEMENS