Software Variability in Service Robots (Summary)

We present “Software Variability in Service Robotics” published in the Journal of Empirical Software Engineering (EMSE) in 2023 [2]. Engineering robotics software systems [1] is a booming discipline, with recent breakthroughs in AI that improve robots’ capabilities to emulate human perception and intelligence, while operating in environments that are potentially unsafe or even hostile to humans. A trend are service robots—autonomous, mobile robots that can address meaningful tasks. For example, consider a disinfection robot for hospital rooms, which was developed during the COVID19 pandemic by one of the subject companies of this paper. Service robotics presents a much more sophisticated challenge for software engineering than traditional, factory-automation type robots, due to several drivers of variability that inherently need to be reflected in software, and lead to variability management issues. In our journal article [2] we survey the state-of-the-art and stateof-practice in software variability in service robotics. We present and triangulate the results from a systematic literature review and an interview suite conducted with eleven practitioners from three subject companies. The paper extends a previous workshop paper [3] that did not yet include the SLR and only two interviews. We now summarize the results for our three research questions. The paper adds details within 38 observations, each paired with actionable recommendations for researchers and practitioners.