Owing to the increasing volumes of systems and software prototyping for electronic equipment, there is a real need to improve software prototyping quality in several major industrial sectors, including the aeronautical and automotive industries, the transport industry as a whole, industrial process control and even consumer electronics. The main obstacle to achieving this quality is the lack of awareness that there is a potential benefit in managing prototypes as important products in themselves, especially in highly constrained domains, where the focus - in terms of development methodology - is given to the final products.
For example, the software activity within the Sextant Avionique Commercial Avionics Division is mainly divided into onboard software and simulation software. The volume of onboard software embedded in airborne equipment has multiplied by a factor of 5 over the last 10 years. Sextant Avionique has acquired a high degree of maturity with this type of real time safety critical software for which methodologies (such as RTSA, Statecharts) and associated tools are of common use.
In comparison, the company's has less maturity with regard to the development of simulation software, especially in so far as concerns the preparation for the future which is of strategic importance in the aeronautical sector. The main reason is that this activity principally concerns prototyping for the demonstration of concepts, without safety constraints and without direct reuse of the code for the airborne equipment.
The objective of PROQUAL is to promote the systematic use of a mature Object Oriented Methodology and the tool that support it to improve the quality of the software prototyping for complex systems, in domains requiring a large and continuous concept definition activity.
The result will be the demonstration of the possibility to reach these two objectives sometimes considered as contradictory:
- the flexibility needed for prototyping complex systems,
- the strict development methodology needed to achieve higher software quality.
The work to be performed is based on both direct and reverse-engineering of part of the baseline project and their comparison with the previously used methods of engineering. It will follow the next main steps:
- initial training in the selected object oriented methodology and associated tool,
- use of a baseline project including two prototypes, Proto A and Proto B
- first step of experiment focusing on reverse engineering of Proto A (code of Proto A is already available) and intra-project reuse of a new version of Proto A (under development)
- systematic evaluation of related quality issues: productivity, generated document and code, knowledge transfer, evolutiveness
- second step of experiment focusing on direct engineering of Proto B (analysis of Proto B is already available) and inter-project reuse between Proto A and Proto B
EXPECTED IMPACT AND EXPERIENCE
The personnel turnover is high in R&D, where a majority of young engineers typically remain for 3 to 5 years, before moving to product development. Furthermore, the crisis that the aeronautical industry is currently going through has led to reduced budgets for R&D activity. It is therefore necessary to improve both the capitalisation on the know-how corresponding to the innovation, and the productivity of the design teams responsible for the R&D activity. This is precisely the aim of PROQUAL. By rationalising prototype software development, PROQUAL will help Sextant Avionique to continue its strategic preparation for the future on a large scale and enable the company to keep up with the global market competition.