Prof. Dr.-Ing. Katja Windt is a member of the Management Board of SMS group, responsible for the divisions SMS digital and Electrical and Automation Systems. In this interview, she speaks about the challenges of the digital future, their impact on the plant and mechanical engineering sectors and, above all, the benefits for the customers.
We want to help our customers get the most out of their machines and processes. Our approach is to combine technology with smart logistics and digitalization solutions that generate added value for our customers. Already today, we offer a range of new products and services based on digital technologies. With these solutions we achieve significant efficiency improvements along the entire supply chain.
Yes, indeed. Digitalization has an impact on strategies, processes, structures and products alike and will lead to sustainable change in the companies taking part in the value added chain. People, machines and commodities will in future communicate in real time. Learning algorithms will transform the value creation chains into dynamically acting value creation networks.
This holds certainly true for the digital technologies and as far as the amounts of data are concerned. But the human operators will receive best possible support in performing their tasks and in decision-making. A key requirement, however, is that the eco-system has a robust and flexible data architecture based on a platform infrastructure. It will integrate different production facilities with different data provision structures. In reality, this will often be a mix of newer and older machines or automation systems. Our experts will combine all this into one consistent system for our customers.
Not only in the future. This is feasible today! We have already implemented various projects with a focus on digital solutions. The learning steel mill, which we have erected in the U.S.A. together with our customer Big River Steel, is an outstanding example of such projects. The smart steel mill continually optimizes the production process from the raw material all the way down to the finished product as part of an integrated supply chain. This is achieved by the use of process know-how and expertise as well as physical and data-controlled models.
Worldwide, we have gained very good and very successful experience. Nevertheless, the task is not an easy one because steel production means that a great many challenges have to be coped with at the same time, all of which have a direct impact on the business result. I’m referring to flexible production planning and production control for varying lot sizes, even single items, while achieving timeliness of delivery and short lead times; to maximum plant performance at minimum maintenance effort and little locked-up capital; to reproducible attainment of best product quality at high yield and without high inventories. All this is taking place under circumstances such as legal requirements, environmental standards, resources, and costs of raw materials, energy and personnel. When we develop a solution, we have to take all these aspects into account - with a view to the specific requirements of the individual customer and within a to-be-defined target conflict corridor.
The target scenario consists of three main elements. First: real-time control of production planning. Second: plant condition monitoring. And third: monitoring and assuring product quality.
All three together form the key to the success of the company. At the same time, they are closely interlinked. Production planning, for example, needs real-time information about how well the production targets are met, that is information about productivity and capacity utilization, the plant condition, the achieved product quality and adherence to delivery dates. This is the only way for the plant operators to ensure compliance with customer specifications. On the other hand, maintenance needs information about the current status and condition of machines and components as well as about resource and capacity requirements for the planned production. This is needed to ensure that the required product quality is actually producible with the current condition of the plant. The product planning requirements are assessed, documented and adapted, if necessary. All this has to be performed in a holistic way.
This becomes obvious and can be experienced by our customers in many ways. Let me just describe one scenario out of many: In a smart steel mill, the centralized, rigid and hierarchical production planning process is being gradually replaced by real-time production optimization. This is a fundamentally different approach because planning becomes an agile process based on real-time data from production. Production planning takes into account any change in circumstances, for example, new orders, imminent maintenance activities or quality deviations, generates alternative scenarios autonomously and evaluates these scenarios. The result is a parameterization on the basis of learning algorithms.
Absolutely. In future, we – and, of course, our customers – will make quality predictions using machine learning techniques and uninterruptedly evaluate the available data for causes of quality deviations. All this will take place in parallel to the running production. In this way, the aimed-for, self-learning quality monitoring processes of digitalization become reality. In this context, our QES/ PQA® system will play a key role in connection with other solutions we provide and with our plant technology.
The plant operator will in future receive best possible support in the performance of his tasks and in decision-making. He will no longer have to search for information in different data file formats or sometimes obsolete hard copies, and interpret the data by himself. Quality is never generated in the last process step, but evolves along the production chain. Here our digital technologies open entirely new perspectives, bringing to light relationships between data and, consequently, between process and product parameters – relationships that were not visible in the past. In this way, we generate new knowledge from data and added value for our customers.