h1

TY  - THES
AU  - Lutz, Marco
TI  - Test suite generation and augmentation for reconfigurable industrial control software in the internet of production
VL  - 2024,09
PB  - RWTH Aachen University
VL  - Dissertation
CY  - Aachen
M1  - RWTH-2024-08524
T2  - Aachener Informatik Berichte
SP  - 1 Online-Ressource : Illustrationen
PY  - 2024
N1  - Weitere Reihe: Informatik. - Weitere Reihe: Technical report / Department of Computer Science
N1  - Dissertation, RWTH Aachen University, 2024
AB  - With the advent of Industry 4.0 and the digitally networked factory, cyber-physical production systems (CPPSs) are reconfigured frequently along their life cycle to adapt to changing customer requirements or market demands. Such reconfigurations are not limited to the hardware but also affect the software of the programmable logic controllers (PLCs) driving these plants. While verification and testing are two techniques capable of alleviating the risk of introducing errors in production code, it is no longer sufficient to rely only on the results obtained by these methods during the commissioning of the CPPS. Even minor incremental reconfigurations to the PLC’s software during the operational phase of the life cycle may introduce regressions that can be quickly overlooked by a developer and therefore need to be reverified. The goal of this thesis is to provide a “push button” analysis for generating test cases after a static reconfiguration. The generated test cases can be injected and monitored during maintenance or virtual commissioning to observe the impact of reconfiguration on the CPPS by the developer. In order to reduce redundancy in test suite generation (TSG) after a structural reconfiguration to the PLC software, symbolic summaries of specific parts of the program should be cached and reused to benefit subsequent analysis. While automatic TSG is an established technique used to generate test suites adhering to structural coverage metrics of PLC software, the generated test suite might not anymore be adequate enough with regards to the coverage metric to ensure the absence of regressions. An indispensable part of regression testing (RT) is test suite augmentation (TSA), which guides the TSG toward the reconfigured behavior and increases the chances of deriving difference-revealing test cases which expose behavioral differences between the program and its reconfigured version. The derivation of new test cases is required to uncover potential regressions after a reconfiguration. To this end, the contributions of this thesis include- heuristics for the scalability of the existing TSG for PLC software,- the reuse of symbolic summaries during TSG of reconfigured PLC software,- and the concept of executing the old and new version of a reconfigured PLC software in one unified program version during TSA. These contributions are evaluated on selected domain-specific benchmarks of varying difficulty from the PLCopen Safety suite and the Pick and Place Unit (PPU).
LB  - PUB:(DE-HGF)11 ; PUB:(DE-HGF)3
DO  - DOI:10.18154/RWTH-2024-08524
UR  - https://publications.rwth-aachen.de/record/993003
ER  -