Five Questions on Digitalization
First of all, we determined what digitalization actually means. We found that the concept was so comprehensive that we decided we would need to focus on implementation. As a matter of the highest priority, we launched into ascertaining what adds value for our customers. This is what we’re principally addressing as most issues surrounding digitalization depend on machines and the sector. They concern what is referred to as “Industrial IT” – which is also the name of a division set up specifically for this purpose. Products are developed that can be used in any machine via a standard interface – regardless of the business unit. An example is our line management system OPAL. Operators of our machines benefit from the digitalization methods employed back in the engineering phase, such as simulation and digital twinning. In the production phase, product data management and product lifecycle management are important tools. And our customer service team uses methods such as remote maintenance and networking with production systems to ensure the maximum availability and performance of machines. Many of these approaches only gained importance along with the “digital revolution” even though they have already been adopted for a while now.
Joachim Dittrich, Optima Consumer
Dr. Georg Pfeifer, Optima Nonwovens
At first, the developments were on a scale that we had not expected. Hardly anyone was talking about “digitalization”; it was significantly more difficult to establish or even maintain networking between customers and manufacturers. Networking had already been around for a while and was used for remote maintenance, for instance. However, there were uncertainties regarding data security in the case of new services, such as cloud servers and cloud computing. When a particularly security-conscious customer – such as a pharmaceuticals manufacturer – hears that hackers obtained access to machine control systems, they are quick to say that they are cutting all connections to the outside world so that the same thing does not happen to them. As a result, networks that were already set up with our customers could no longer be used as they were previously.
Nevertheless, digital twins and networked SCADA systems continue to be created. However, in most cases this takes place on strictly disconnected server systems (on premise) and not on outsourced systems to which the manufacturer and operator have equal access. Up to now, the question of who data belongs to was of secondary importance; now however, this question plays a much greater role. Information – referred to as the gold of the 21st century – has been brought to the fore with digitalization and networking.
This especially applies to large companies that approach everything to do with digitalization in a very structured way. With smaller companies, it’s often simply a question of displaying data. In contrast, we are seeing an almost scientific approach being adopted by some major companies. They are also interested in standardization: If efficiency has been increased at a site by means of adopting digitalization measures and this has resulted in improvements, this approach is also transferred to other factories.
The umbrella term “Optima Total Care” refers to all of our products that are principally focused on supporting our customers and their processes. This support may be needed in a whole host of different areas. It all begins with a better understanding during the design phase with simulations, digital twins and virtual reality. It goes beyond commissioning with the provision of data collection tools, process data simulations and reliability evaluations. And it also concerns production with production planning tools, condition monitoring and digitally supported maintenance, such as with spare part and purchasing systems and planning tools.
It is already possible to collect very detailed data from process steps and statuses, with a huge volume of data being amassed within milliseconds. However, “big data” is not yet leading to the results expected today. First of all, it is necessary to manage the flood of data. In any case, some machine operators already have considerably more insight, being able to see, for example, how the motor current of their machine behaves over time. After all, the vast volumes of incoming data can only be exchanged between the operators and machine manufacturers with difficulty. Terabytes of data can only be used in a meaningful way by intelligently compressing the volumes of data via self-learning algorithms implemented by artificial intelligence systems.
As in the past, the technician casually leans against the machine – with one hand on the machine frame – and observes the process. The technician hears every noise and detects varying vibration levels with their eyes and hands. That is data collection par excellence – as well as evaluation – as the experienced technician identifies the damaged bracket – to be replaced soon – without any problems in the majority of cases. In the future, this will be the task of condition monitoring and predictive maintenance – a task that is anything but easy to master. And it’s often not worth analyzing the parts for which issues are easy to foresee – perhaps a defective pneumatic cylinder – as the respective information is exchanged within just a few minutes. Analyzing big data often has limitations too, such as when it is ultimately only possible to find correlations instead of causalities. It is more advantageous to introduce the information in certain contexts, such as when it is observed that a particular fault occurs more frequently following specific operator actions. It is then possible to retrace the cause of the fault. Developing a rule on the basis of this and then teaching the machine how to remedy issues is then a whole other challenge. The first step for artificial intelligence is for the machine to learn to identify recurring problems itself and automatically react as an operator would.
It’s very often a matter of ensuring more efficient production, which we already achieve today via appropriate data evaluation. Just being aware of what is influencing the OEE by means of evaluating the data with the OPAL monitor increases production efficiency in most cases, as focused measures can be adopted. The transparency and availability of operating and process data enable much more targeted actions to be taken today than were previously possible by just noting the information.
Virtual reality, augmented reality or mixed reality can already be used during the system design and planning stages, as well as during training and will support maintenance and servicing operations in the future. These tools are also opening up new opportunities to support personnel directly with regard to machine operation – user experience being the key term here – or modifications. Our objective is for any operator to be able to operate our machine solutions in an optimal way – regardless of their education, background or language.
And this will only be possible if we keep our ears to the ground at all times when it comes to machinery – or in other words, have access to industrial IT solutions. What we understand as digitalization today is very different to how it was seen in the past: an almost unconnected discipline independent of the machine type. Industrial IT offers added value that goes beyond machine boundaries by means of evaluating data via networking solutions. As such, our employees’ qualifications also need to change. Expertise in the fields of IT, web programming, data bank structures and cloud services are becoming more and more necessary. These experts are working together in cross-divisional structures. As ever, our objective is to enhance customer benefit.