In this project we focus on the specific challenges of automated testing for real-time, embedded systems.
In particular, we consider model-based testing, in which models of both the system and its context (i.e., the environment in which it operates) are employed.
To achieve its objectives the project will improve, automate and optimize large parts of the testing process by developing prototype tools (Eclipse plug-ins) backed by novel techniques for:
- improved system level testability through context models capturing all relevant system requirements and extra-functional specifications for timing and performance, for example. Context models are abstractions of the environment in which the system under test will execute. They enable simulation of the environment and automated testing of the system under test.
- automated testing at the early stages of a project by means of
- automated test case generation from context models
- automated oracle generation from context models
- automated test execution in a simulated environment, on the development platform
Novel automated test techniques such as Adaptive Random Testing and search-based testing will be investigated, improved, and tailored to you specific objectives. Based on industrial requirements from real cases we will validate the novel techniques and the tool prototypes in industrial cases provided by our industrial partners.
Partners:
WesternGeco, TOMRA, ABB, Sintef
