NEWS
-
Shift to model-based software testing
The rapid growth in complexity of embedded software has raised serious
quality concerns, particularly in safety critical applications, such as
automotive, military, space and medical systems. In addition to a
structured, even formal, design approach, the focus now is on test and
verification. Model based design has become the preferred
approach for many of these complex applications, because it improves
the specification, design and implementation phases. The dynamic
trend today, however, is the shift to model-based testing.
New tools are fast emerging for model-based design. Techniques such as developing executable specifications and automatic code generation are not only helping embedded software development, but are crucial to addressing the test and verification challenges, and thereby improving embedded software quality.
Next generation software development and test tools must extend the model-based design approach to include requirements-based testing, coverage analysis, and test generation. They should be flexible to enable either a strict top-down constrained approach, or an unconstrained bottom-up design that can be constrained later, to ensure software is more robust and testable.
Designers must rely on these tools not only to facilitate model checking and test generation, but also for model and test data reuse at different levels and stages throughout the development process. Another important aspect will be the ability to manage the huge amounts of data generated as part of the development of complex embedded software systems, such as model diagrams, state charts, test cases, test benches and test results.
The tools have a lot to live up to. The importance of managing model-driven design combined with model-, rather than code-driven testing, is the ability to trace test cases back to use cases. Use case analysis is becoming critical for efficient and effective verification of the most complex systems.
New tools are fast emerging for model-based design. Techniques such as developing executable specifications and automatic code generation are not only helping embedded software development, but are crucial to addressing the test and verification challenges, and thereby improving embedded software quality.
Next generation software development and test tools must extend the model-based design approach to include requirements-based testing, coverage analysis, and test generation. They should be flexible to enable either a strict top-down constrained approach, or an unconstrained bottom-up design that can be constrained later, to ensure software is more robust and testable.
Designers must rely on these tools not only to facilitate model checking and test generation, but also for model and test data reuse at different levels and stages throughout the development process. Another important aspect will be the ability to manage the huge amounts of data generated as part of the development of complex embedded software systems, such as model diagrams, state charts, test cases, test benches and test results.
The tools have a lot to live up to. The importance of managing model-driven design combined with model-, rather than code-driven testing, is the ability to trace test cases back to use cases. Use case analysis is becoming critical for efficient and effective verification of the most complex systems.
More articles in this category:
 | |||
| |||
 |
PRODUCT INNOVATIONS |
|
|
| |||
| |||
| |||
| |||
 |
ANALYSIS-MARKET-TRENDS |
|
|
| |||
| |||
| |||
| |||
|
NEWS |
|
|
| |||
| |||
back
top
