Düsseldorf, August 29, 2017

The vision of a "learning steelworks"

Everywhere in the world, steel producers are presented with similar challenges: First comes the customer, for whom everything possible has to be done to ensure that he receives his product in the specified quality at the agreed delivery date. At the same time, optimal use has to be made of the existing capacities, while minimizing maintenance activities and locked-up capital. With the number of small batches and special orders increasing and timeframes becoming ever tighter, future production planning will have to be more flexible, precise and – ideally – realtime- capable.

For all this, digitalization is an indispensible requirement. The “smart steelworks” relies on the assumption that data and KPIs derived from such data are recognized and accepted as a basis on which business and processrelevant decisions are taken. Digitalization primarily entails intelligent merging of knowledge and data and the systematic networking of the two. Today, knowledge and data are still very often unsystematically spread over the company organization or stored in data silos with strict rules of access. What needs to be done is to overcome exactly this way of thinking and make information available in real time - both internally, within the company, and partly also externally, across company boundaries – and use the data as a decision-making basis.

This may sound simple. But in actual life it may turn out to be a real challenge. Quite often there is not a lack of data, but a lack of data quality and data consistency. Often also an appropriate data aggregation process or simply the correlation of data to the product or its localization are missing. In the “smart steelworks”, production and planning data as well as plant and product data have to be made available in real time in a Unique Data Source of Truth, which the central operative management systems will use as the basis for production planning, product quality and maintenance processes. In the “smart steelworks of the future”, the still common centralistic, rigid and medium- to long-range production planning practice will be replaced by real-time selfoptimizing production planning. Every incident (a new order, a modification of the maintenance schedule, a quality deviation, etc) will lead to a review of the most recent plan, i.e. the system will check in real time and in line with clearly defined key performance indicators whether the last planning result could possibly be replaced by a better one. In doing so, the system will take into consideration - in real time - the data content of the Unique Data Source of Truth and adopt a structural production process model.

Software facilitates work

The second challenge is to document and track the quality of the products along the steelworks’ entire value creation chain. Immense quantities of process and product data arise at each one of the numerous production stages. All the data has to be set in proper relation to the products. Provision has to be made for the fact that slabs and strips do not only change in length, but may also have to be side trimmed and/or turned. Alone a software solution making these data transparent and intuitively accessible in real time would be a great support for the operating and quality control staff. They are actually the ones responsible for taking care that the process runs properly and the customers will not be displeased with the quality of the products they receive.

Also quality decisions have to be taken in real time in order for the operators to be able to take immediate action, for example, by intervening in the process or assigning products to another customer order. For this, modern software solutions, such as the Product Quality Analyzer – PQA, operate with sets of rules, checking the actual condition of every product within the process chain against the specified product and customer requirements.

A third challenging field of activities is to capture and map the current conditions of the plants and components and keep track of the respective maintenance requirements within a Computerized Maintenance Management System (CMMS). Designed as an Asset Management system, such system first aggregates all relevant information (item lists, drawings, documentation, data sheets, etc) and the corresponding maintenance instructions. Via a link with the ERP system and the warehouse management systems, the maintenance personnel can identify, request or, if necessary, purchase any required items or components. The CMMS triggers the relevant maintenance orders and documents their accomplishment in digital form. Additionally, data-based evaluation routines provide maintenance-relevant information or predictions about the condition of plant components. The data for such routines may come directly from the automation system or may be provided by preventive maintenance systems, which often operate with artificial intelligence, or by condition monitoring systems in place at the individual components. Especially the latter functions can be expected to be increasingly performed by the plant suppliers as a service product in the future.

For more information please read the detailed version of this article published in the SMS group Newsletter 01/2017.