Abstract
After a number of success stories in safety-critical domains, we are starting
to witness applications of formal methods in contemporary systems and software
engineering. However, one thing that is still missing is the evaluation
criteria that help software practitioners choose the right formal method for
the problem at hand. In this paper, we present the criteria for evaluating and
comparing different formal methods. The criteria were chosen through a literature
review, discussions with experts from academia and practitioners from
industry, and decade-long personal experience with the application of formal
methods in industrial and academic projects. The criteria were then evaluated
on severalmodel-oriented state-based formalmethods. Our research shows that
besides technical grounds (eg, modeling capabilities and supported development
phases), formal methods should also be evaluated from social and industrial
perspectives. We also found out that it is not possible to generate a matrix that
renders the selection of the right formalmethod an automatic process. However,
we can generate several pointers, which make this selection process a lot less
cumbersome.
Original language | English |
---|---|
Number of pages | 30 |
Journal | Software: Practice and Experience |
DOIs | |
Publication status | Published - 2018 |
Fields of science
- 102 Computer Sciences
- 102022 Software development
JKU Focus areas
- Computation in Informatics and Mathematics
- Engineering and Natural Sciences (in general)