Contact Person
Prof. Dr. Ina Schieferdecker
Prof. Dr. Ina Schieferdecker
Director FOKUS
Tel.: +49 30 3463-7241

D-MINT

Apr. 01, 2007 to Dec. 31, 2009

There are more and more possible applications for software-intense systems. The requirements on such systems increase and that enlarges their complexity. That is why, it is essential to develop such technologies within short time and with decreasing error-proneness.

The D-MINT consortium will develop with 27 partners a new approach to model-based testing which combines various system aspects and system models during test generation. Every partner provides branch-specific requirements, concepts and ideas. The results are combined in a technology for automated software-engineering.

On the following sites you can learn more about current challenges of software-engineering, the main goals of D-MINT and the deployed technologies. When you are interested in this or related topics please contact us.

Award

The ITEA2 project D-MINT on Deployment of Model-Based Technologies to Industrial Testing won the ITEA Exhibition Award 2009. FOKUS contributed to D-Mint as research Partner and coordinator of the German consortium.

During the Final Project Review Meeting on 24.11.2009 in Espoo, Finland, the project was rated Excellent.

Goals

The following major results are expected from this project:

  • Packaged methods and techniques for model-based testing processes to enable comparable developments beyond the selected technologies and tools of this project including guidelines and experience reports derived from the case studies.
  • Industrial demonstrators showing the practical and efficient use of model-based testing resulting from case studies in automotive, finance, industrial automation, manufacturing equipment, railways, and telecommunication.
  • Tool chain for test generation, test refinement, distributed test process control and test quality analysis including new and extended tool components for commercial tools from partners and/or for off-the-shelf tools.
  • Case study experiences and evaluation results that prove the efficiency of the developed processes and methods empirically
  • New algorithms for architecture-driven test modelling and test generation with industrial scale
  • Extended set of test patterns for software-intense systems in the domains of automotive, finance, industrial automation, manufacturing equipment, railways, and telecommunication
  • An exploitation plan in yearly revisions. It will be carefully planned and starts at the beginning of the project. This includes analysis of market potential and focusing/defining the market with regards to the vertical, horizontal, regional, and company-size dimensions.

Technology

The D-MINT innovations will be in the following five key areas:

1. System-architecture-driven testing
A new approach for model-based testing will be acquired where the system architecture serves as an integrator of heterogeneous component models and where test patterns serve as a guide for test generation along industrial expertise. The architecture-oriented model allows us to combine aspects, relations and models of different kinds of components.

2. Model-based integrated system and test development
There will be developed concepts for the management of requirements and test-cases consistency for distributed and compositional system structures. Models on different levels and from different perspectives including architectural, hierarchical models will help us to manage system development and tests for systems consisting of huge amounts of heterogeneous components.

3. Automated test-case refinement in sync with system model refinements
We will develop an automatic concept to let test-cases keep pace with the refinement of models during design. We will model transformers that refine the tests. Furthermore, we will give emphasis to the practicability of generated tests by investigating means to steer test generation with proven test patterns and the ability to control test generation via explicit user interaction.

4. Automated consistency checks of requirements, models and test cases
We will address consistency by keeping track of the relations between requirements, system models and test models (and parts thereof) on a semantic and not just syntactic base. This means that not only the sole fact that model elements relate to each other, but also the transformers being applied are traced as well as the history of element evolution.

5. Statistically controlled model-based test processes
Statistical methods should be applied to model-based test processes to measure, control, and optimize process and system qualities. Therefore, D-MINT will develop concepts, methods, and tools for requirements formalization towards probabilistic models, probabilistic test modelling, test-case generation (from probabilistic models of system use and behaviour), and statistical methods for analysis of test results that apply to domain-specific composed software-intensive systems.

Partners

  • iXtronics
  • PikeTec
  • Testing Technologies
  • Fraunhofer IESE
  • DaimlerChrysler, ABB und Nokia
  • INSPIRE AG

More information

With the ever increasing penetration of software-intense systems into technical, business and social areas, not only the requirements on system functionalities and features, but also the requirements on system quality and reliability increase. With the increasing requirements, the complexity of software-intense systems is growing – combined with an increasing error-proneness which is also related to shortened development times.

To remain competitive capable and cost-efficient technologies for development and testing are required. The currently existing technologies have still many deficits:

  • Current model-based test approaches lack industrial scale and can not be used in full range of industrial practice.
  • They are also limited to selected aspects of system models.
  • A typical software-intense system is developed as a composition of software components, electrical components, and/or mechanical components each of which uses different modelling techniques.
  • Because of different deployed modelling languages the individual system components are not consistent to each other.
  • Development of test technologies causes great costs.
  • Development is time-consuming.

Model-based testing technologies offer the potential for a high degree of automation to increase effectiveness and efficiency because test cases can be derived mechanically.
D-MINT will provide partners and European industry with a new leading edge model-based testing technology that enables the testing of software-intense technical systems and the assurance of their quality. These technologies allow the production of high quality software-intense systems at reduced expenditure in terms of time and money.