Location

Room E.0.04, Building E
Campus Etterbeek, Vrije Universiteit Brussel
Pleinlaan 2, 1050 Brussel

12:30 - 13:00

Welcome lunch with sandwich

13:00 - 16:30

Advisory board meeting and presentations:

13:00 - 13:15 – Welcome
13:15 - 14:00 – Keynote
"Fuzzing Web APIs in Industry with EvoMaster"
Prof. Andrea Arcuri
(Abstract and Bio below)

14:00 - 16:00 – Status of the proof-of-concept pilot projects

Resilience Testing
Failure Reproduction (Piloted at KatoenNatie)
** Coffee break **
Flakey Tests (Piloted at Nokia)
Test Amplification (Piloted at KatoenNatie)

16:00 - Closing of the meeting –
Quest for additional pilot projects next year

16:30 - ...

Reception with networking drink

Keynote: "Fuzzing Web APIs in Industry with EvoMaster"

Abstract:
As researchers in academia, we have worked for the last 9 years on the open-source fuzz-testing tool called EvoMaster. A fuzzer is able to automatically generate test cases, automatically finding different types of faults. In the case of EvoMaster, we can automatically generate test cases for REST APIs (as well as GraphQL and RPC), with output test suites in different formats such as Python or Java. Faults that are found are based for example on HTTP 500 status code (server error), mismatches between the API responses and what declared in their OpenAPI schema, and different kinds of security related properties that are not satisfied.

In our work, we have been pushing forward the state-of-the-art in AI research for software testing. But are those techniques actually useful for test practitioners? What important aspects are missing or neglected?

In this talk, we will discuss in more details our collaboration/engagement with test engineers at Fortune 500 enterprises such as Volkswagen in Germany and Meituan in China. We share the experience on how testers have been using this kind of tool to simplify their work.
Anyone can download and use an open-source tool such as EvoMaster. However, industry-academia collaborations are important for academics to get access to real-world systems, whereas the practitioners benefit from feature-requests/bug-fixing prioritization.

Short bio:
Prof. Andrea Arcuri is a Professor of Software Engineering at Kristiania University College and Oslo Metropolitan University, Oslo, Norway.

His main research interests are in software testing, especially test case generation using AI techniques. Having worked 5 years in industry as a senior engineer, a main focus of his research is to design novel research solutions that can actually be used in practice. Prof. Arcuri is the main-author of EvoMaster and a co-author of EvoSuite, which are open-source tools that can automatically generate test cases. He received his PhD in search-based software testing from the University of Birmingham, UK, in 2009.