Mazkatli, ManarArmbruster, MartinKoziolek, AnneHerrmann, Andrea2023-11-302023-11-3020230720-8928https://dl.gi.de/handle/20.500.12116/43232Architecture-level performance models (aPMs) provide valuable insights for design decisions with performance predictions. Due to source code changes and system adaptations, it remains challenging to keep aPMs up-to-date and ensure their consistency during agile software development. The Continuous Integration of Performance Models (CIPM) approach focuses on maintaining the consistency between aPMs and software artifacts. After each significant change during the software development and operation, CIPM automatically updates aPMs. However, the current implementation of CIPM is limited to Java- and microservice-based applications. In this paper, we evaluate whether conceptual changes are required if CIPM is applied to the Lua programming language and industrial sensor applications. Our evaluation is based on a real Lua-based sensor application from the SICK AppSpace ecosys tem and an artificial one. The findings demonstrate the feasibility of CIPM as no significant conceptual changes were required, but rather technical ones.enSPPperformancepredictionarchitecture-level performance modelContinuous Integrationsensor applicationText/Conference Paper