Automated Software Testing

Automated Software Testing

Software Engineering for data-intensive applications

Automated Software Testing

The Automated Software Testing field of competence includes research into the latest testing techniques to improve the reliability and dependability of data-intensive software systems, including deep neural network-driven cyber-physical systems (CPS) such as autonomous vehicles or warehouse and delivery robots.

The researchers develop new tools and frameworks for automated testing, improve existing testing methods, and explore new ways to integrate testing into the software development lifecycle. For example, mutation testing is used to evaluate the quality of existing software tests and fuzz testing is adopted to produce unexpected data that assess the correctness of a computer program under corner-case conditions.

fortiss advances the state-of-the-art in the field of software testing, focusing both on improving the efficiency of the testing process and enhancing the accuracy and completeness of the testing results. To this end, the researchers collaborate with national and international industry partners to apply their research results in real-world settings. Please contact us, if you are also interested in collaborating with us.

 

The aim is to enhance the quality and reliability of software systems by 

  • automating the testing process,
  • reducing the need for manual testing and
  • improving the overall efficiency of software development. 

In doing so, the researchers target the quality of software systems to reduce the likelihood of software failures in data-intensive applications, including web applications and AI-enabled cyber-physical systems (CPS) such as autonomous vehicles.

 

    Research focus

    The current main research areas include

    • Post-production testing in which we deal with methods to ensure the high dependability of CPS in production. Investigated techniques include run-time monitoring and supervision using uncertainty quantification and reasoning to provide fail-safe mechanisms and self-healing capabilities. Moreover, we investigate how to perform regression testing of highly-adaptive systems (e.g., those based on continual or reinforcement Learning).
       
    • Testing collaborative CPS in which we investigate how to test collaborative autonomous and mission-critical CPS that achieve a unified goal. Current challenges are related to the notions of software defect and software oracle, which need to be revisited. Potential use cases are related to autonomous driving systems, disaster relief systems, autonomous agricultural machines and warehouse and delivery robots.
       
    • AI-enhanced testing in which we leverage the most powerful AI toolsets to make testing more effective and more efficient. We use generative adversarial techniques, conformal predictions, explainable AI for testing, to fill the gap between simulated and real-world platforms for CPS. Uses cases are related to customized/personalized testing platforms and digital twins, hybrid simulation techniques that leverage both real-world data and neural techniques to improve the fidelity and trustworthiness of simulation-based testing. Our framework considers both simulation-based testing and the transferability of test results from simulated to real platforms.

    By activating this video, you consent to transmitting data to YouTube.

    Practical Course

    Advances in Testing Automated Driving Systems

    Together with the Technical University Munich (TUM), the fortiss competence field Automated Software Testing offers a brand new practical course for the upcoming winter semester. The course is centered around the topic of autonomous driving testing.

    Read more

    Prof. Dr. Andrea Stocco

    Your contact

    Prof. Dr. Andrea Stocco

    +49 89 3603522 271
    stocco@fortiss.org

    Publications

    • 2024 OpenSBT: A Modular Framework for Search-based Testing of Automated Driving Systems Lev Sorokin, Tiziano Munaro, Safin Damir, Brian Hsuan-Cheng Liao and Adam Molin volume , , , edition, 2024. , . Details URL DOI BIB
    • 2023 A Retrospective Analysis of Grey Literature for AI-Supported Test Automation Filippo Ricca, Alessandro Marchetto and Andrea Stocco In In Proceedings of the 16th International Conference on the Quality of Information and Communications Technology, pages 90–105, Springer Nature Switzerland. Details URL DOI BIB