Industrial Automation & Control Systems are key elements in any production system for both discrete manufacturing as well as for batch and continuous production processes.
One challenge nowadays is that the cost for programming and commissioning of manufacturing systems is covering the largest fraction of cost for new manufacturing systems. The lack of re-usability and re-configurability of industrial control systems software requires a large effort for re-programming the production system at every system, process, or hardware change. Moreover, the enhanced degree of vertical integration into MES and ERP systems, horizontal integration, interaction of sub-plants and whole production systems, and the amount of collectable manufacturing data offer complete new opportunities and business models for future manufacturing, but also pose new challenges, commonly known under the label of “Industry 4.0”.
The goal of our research in this field is to simplify the engineering process of industrial automation systems in general, and the industrial control software engineering in particular. We develop automatic and semi-automatic code-generation approaches for industrial manufacturing systems. Based on extracted knowledge of existing engineering data from planning documents, a full model of the plant is generated, which enables the linking of its control program, system documentation, and re-configuration.
To handle the complexity of larger scale systems and to systematically address modular automation systems, our research targets the development of methods towards a modular system architecture that encapsulate individual automation components in hardware as well as in software, which provide particular services (or skills) to the overall automation system. This modular approach enhances the re-usability of complete sub-systems, e.g. process tanks, as well as re-configurability when replacing single components, such as e.g. a pneumatic actuator by a linear motor, without the need for re-programming the manufacturing system.
Another research focus for modular automation systems is the efficient integration of modern control methods, such as self-learning approaches or model-predictive control (MPC), directly onto programmable logic controllers (PLC) for industrial control, which is important for a myriad of typical control applications, like high-speed pick-and-place systems, optimal control of heating systems, or energy management in smart grids.
Our research on modular automation systems is complemented by the development of modular robotic systems as a potential application domain. Modular and mobile robotic systems are developed for a variety of industrial applications, e.g. for automation in civil engineering applications at construction sites.