Design rules for the transformation of automation software architectures to CPPS (Trafo4A@AT-SW)

• Applicant: Professorin Dr.-Ing. Birgit Vogel-Heuser
• Subject Area: Software Engineering and Programming Languages ; Automation, Mechatronics, Control Systems, Intelligent Technical Systems, Robotics
• Term: since 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 548391107

To achieve the highest-possible productivity and efficiency in industrial production, future Cyber-Physical Production Systems (CPPS) must be (semi-) automatically adaptable over several decades, e.g. to new customer requirements (especially relevant in special purpose machinery and plant engineering) or laws and regulations. Since automation (AT) software is typically easier to adapt cost efficiently and on short notice than AT hardware, changes are preferably realized in software. Thereby, the change realization is easier if software is already modular and configurable. Architectures are required that support modularity which sensibly complies with hardware and behavior in AT software. More than 50% of AT software realizes diagnosis and fault handling, operation mode changes, and the ramping up or down of the CPPS. These extra-functional tasks are essential for the development of an architecture, i.e. the sum of all design decisions regarding quality attributes like reusability or configurability. However, such complex dependencies have barely been researched to date and are a focus of Trafo4A@AT-SW. The aim of Trafo4A@AT-SW is therefore the transformation of existing legacy AT software and the respective architectures of special purpose machine and plant manufacturers towards CPPS-capable AT software architectures. Automatic static and dynamic analysis of IEC 61131-3 based AT software quality serves for the identification of potentials for modularization and quality improvement in existing architectures. The categorization of existing AT architectures, the research into industrially feasible design rules for them, and the identification of a suitable stepwise combination of those rules are methods to achieve the goal. Thereby, boundary conditions of the CPPS manufacturer will be included in the models and the software architecture will be iteratively transformed into the desired CPPS architecture. The solution concept addresses three partial goals: first, the development of categories to assess AT software architectures, second, the deductive derivation of adaptable design rules, and third, methods to identify transformation trajectories for the architectural transformation. To reach that goal, two companies in special purpose machinery and plant engineering and one AT platform provider (cf. LOIs) are available to iteratively evaluate requirements and partial results as well as for the comprehensive evaluation using the prototype that will be created. This prototype builds on previous work of the applicant to assess software quality. The results are reproduced independently from company-specific applications on the institute’s lab demonstrators to make them accessible to the research community.

• DFG Programme: Research Grants