This abstract summarizes the paper accepted at the Journal of Systems and Software with the same title [1].
Automated production systems form the backbone of the world’s industrial production. They are highly specialized technical systems, which are comprised of mechanical, electrical and electronic parts and software, all closely interwoven. Software is the defining factor to realize modern trends in manufacturing as defined by mass customization, small lot sizes, high variability of product types, and a changing product portfolio during the lifecycle of an automated production system. However, the software can control the production system only via electrical and mechanical means, which have their individual life cycles and undergo evolution. Hence, the evolution of automated production systems always requires addressing cross-disciplinary evolution challenges.
We provide an interdisciplinary survey on the challenges, the state-of-the-art as well as research directions with respect to the evolution of software in automated production systems. In the first part, we discuss the specifics of the development process as well as the different types of evolutions during the system’s life-cycle. In the second part, we survey the challenges due to evolution covering the different development phases (requirements, design, implementation, validation & verification) and several important cross-cutting aspects (variability management, model-driven engineering, and traceability), and review existing potential approaches. Finally, we outline future research directions for the discussed challenges.