Flexible Variability Mining for IEC 61499 Control Software Applications with Varflix
Alexander Stummer, Anna-Lena Hager, Rick Rabiser · 2025
Cyber-physical production systems (CPPS) are industrial automation systems that combine physical components, such as machines, robots, and sensors, with digital components, such as software, networks, and databases.They are made up of heterogeneous artifact types and are often maintained for decades.During their lifetime, many different variants are created, typically following Clone & Own development practices.However, even though such systems would be well suited for applying Software Product Line (SPL) techniques to achieve more systematic reuse, this requires re-engineering these systems.Doing this manually is infeasible, especially in industrial practice.Automatically mining variability information from existing variants is thus desirable.In existing automatic mining approaches, the quality of the results is often not ideal due to a lack of considering domain specifics.Also, these approaches often focus on a particular programming language or artifact type.Flexibility of the mining approach is thus another key factor due to the heterogeneous nature of systems such as CPPSs and the artifacts they are made up of.In this paper, we present our approach for mining variability from CPPS control software developed using the domain-specific standard IEC 61499.This approach combines automatic processes with manual user involvement to better integrate domain knowledge into the process.Furthermore, we provide support for heterogeneous artifacts.