Periodic Reporting for period 1 - SYNTACT (Reactive Synthesis in Action)
Okres sprawozdawczy: 2022-05-01 do 2023-10-31
Bugs in software systems pose a huge and growing cost on society, not only in economic terms but also in direct risk to human lives. Reactive synthesis is a new technology to automatically generate correct-by-construction software. Unlike testing and formal verification, which require advance investment in an implementation, reactive synthesis generates a correct implementation directly from its formal specification.
Despite its potential to reduce software development costs and at the same time increase the confidence one may have in the correctness of software, to date, no application of reactive synthesis has been examined outside academic labs. The main objectives of the project are to put reactive synthesis technologies to test in industrial beta-sites and thus generate and disseminate knowhow on the use of reactive synthesis in practice, identify and address product and process obstacles, build a community of parties that are interested in the promotion and transfer of reactive synthesis technologies from academia to industry, and lay a foundation to guide future research, standardization, collaboration, and industry investment in reactive synthesis technologies.
All these are expected to advance the state-of-the-practice in reactive synthesis and thus start paving the way for its adoption in industry.
Despite its potential to reduce software development costs and at the same time increase the confidence one may have in the correctness of software, to date, no application of reactive synthesis has been examined outside academic labs. The main objectives of the project are to put reactive synthesis technologies to test in industrial beta-sites and thus generate and disseminate knowhow on the use of reactive synthesis in practice, identify and address product and process obstacles, build a community of parties that are interested in the promotion and transfer of reactive synthesis technologies from academia to industry, and lay a foundation to guide future research, standardization, collaboration, and industry investment in reactive synthesis technologies.
All these are expected to advance the state-of-the-practice in reactive synthesis and thus start paving the way for its adoption in industry.
We list the project's main technical and scientific activities and achievements.
First, the study of the Spectra specification language, including extensions of usability and case study examples, based on discussions with collaborators in industry and feedback from early end-users. These include the extension of Spectra with additional language constructs together with user studies that focused on examining the usage of these constructs by end-users.
Second, extending the Spectra synthesis IDE, based on feedback from end-users, with improved performance for important analyses as well as new mechanisms for reuse and evolution.
Third, running several case-studies with industrial partners, collecting feedback from engineers who used the synthesizer. The feedback was the source for the identification of the gaps remaining on the way for larger adoption of synthesis technologies in industry.
First, the study of the Spectra specification language, including extensions of usability and case study examples, based on discussions with collaborators in industry and feedback from early end-users. These include the extension of Spectra with additional language constructs together with user studies that focused on examining the usage of these constructs by end-users.
Second, extending the Spectra synthesis IDE, based on feedback from end-users, with improved performance for important analyses as well as new mechanisms for reuse and evolution.
Third, running several case-studies with industrial partners, collecting feedback from engineers who used the synthesizer. The feedback was the source for the identification of the gaps remaining on the way for larger adoption of synthesis technologies in industry.
We have identified that the key gaps remaining on the way for larger adoption of synthesis technologies in industry are in the ability of engineers to write correct formal specifications, which in turn depends on the specification language, the process and methodology of writing and evolving specifications, and the tools that support them. Further collection of evidence from end-users in industrial settings as well as much further research in specification comprehension, methodology, and tools development, are required in order to start making the use of synthesis in industry a reality.