Automotive software product lines for ECU software configuration: A systematic literature review
Yannick Lindebauer, Richard von Esebeck, Thomas Vietor · Journal of Systems and Software · 2025
In the automotive industry, vehicles are increasingly configured by software to accommodate individual customer preferences. This leads to a growing number of software variants and calibration parameters that must be managed consistently. Efficient handling of this variability benefits from the application of (systems and) software product line engineering (SPLE), as it offers structured mechanisms for reuse, traceability, and systematic variability management. These benefits are particularly relevant for electronic control unit (ECU) configuration, where maintaining consistency across features, parameters, and vehicle variants is crucial. However, existing approaches in industry, e.g. those relying on configurable bills of materials, indicate that SPLE has yet to see widespread adoption in the configuration of ECUs. Instead, proprietary approaches dominate, often lacking integration, transparency, and scalability. This study investigates the reasons why SPLE is not widely adopted for this use case. The underlying hypothesis suggests that industrial requirements are not accurately captured in academic research, making the implementation of scientific SPLE concepts difficult. To examine this assumption, a systematic literature review was conducted, analyzing relevant publications. The findings indicate a pressing need for closer collaboration between industry and academia to better identify challenges and requirements. Furthermore, current and emerging developments, such as software-defined vehicles (SDVs), require greater consideration in ECU configuration research. Our hypothesis was largely confirmed, indicating that SPLE research must be further extended and refined to meet practical ECU configuration needs. Accordingly, a concise, end-to-end methodology is needed to support SPLE-based calibration processes in SDV environments with increasingly decoupled hardware and software.