Skip to main content
European Commission logo print header

Design for validation

Cel

The DeVa project aims at providing methods and tools for validating critical computing systems. DeVa will place emphasis on software validation mainly with respect to dependability requirements rather than functional requirements, and will concentrate particularly on issues of software structuring that will aid the design for validation of real-time distributed systems.

This will be achieved by a programme of long term research on a set of closely inter-related topics concerning system specification, design, implementation, verification and evaluation. DeVa will work mainly on problems and potential solutions of general applicability, rather than concentrating on a particular industry sector, or on particular types of computing system (e.g. hard versus soft real-time systems), or on particular attributes of dependability. Such work will of course be very usefully complemented by projects that are being undertaken by other researchers addressing more specific problems and circumstances, e.g. within specific industry sectors.

The set of closely inter-related results that DeVa plans to provide demonstrations of, and to document in published scientific papers, concern the three main aspects of design for validation, namely understandability, verifiability, and evaluatability. With respect to understandability, DeVa plans to use advanced object-oriented design techniques to deal with logical complexity and tackle dependability issues. For example, techniques such as reflection and delegation will be used to implement adaptive behaviour, and a software architecture based on the use of idealised fault-tolerant components will be used to achieve fault tolerance, and facilitate component re-use. With respect to verifiability and evaluatability, the work will be aimed at enriching this design paradigm in order to turn it into an effective "design for validation" paradigm, e.g. via work on (i) disciplined approaches that ensure the testability of the ensuing design, and (ii) evaluations of software architecture trade-offs that encompass the reuse of existing components. An overall project result which it is hoped will prove to be of great industrial relevance will be the production of a set of design guidelines and assessment techniques to simplify and support the validation and certification of dependable systems.

DeVa's plans are in part based on the work of the ESPRIT Basic Research Actions on Predictably Dependable Computing Systems (PDCS and PDCS2). Four of the original members of PDCS and PDCS2 (City, LAAS, Newcastle and York), whose expertise centres on object-oriented system structuring, system validation and evaluation, and real-time systems, will be joined in DeVa by (i) Universität Ulm, working on formal verification of software, (ii) the UK Defence Research Agency (DRA), working on system security and the problems of structuring and validating the safety and security of distributed systems built from commercial software components, and (iii) the Ecole Polytechnique Fédérale de Lausanne (EPFL), working on specification-based testing methods of object-oriented systems.

In addition to these seven project members, there will be three subcontractors involved in DeVa, namely the AIB-Vincotte Nucleaire working on certification and licensing of safety-critical systems, Université Paris-Sud working on algebraic specification and testing, and the Technische Universität Wien working on time-triggered hard real-time systems. The guidance of an Industrial Advisory Board will help to ensure industrial relevance and take-up of DeVa's results.

Zaproszenie do składania wniosków

Data not available

System finansowania

CSC - Cost-sharing contracts

Koordynator

UNIVERSITY OF NEWCASTLE UPON TYNE
Wkład UE
Brak danych
Adres
6 KENSINGTON TERRACE
NE1 7RU NEWCASTLE UPON TYNE
Zjednoczone Królestwo

Zobacz na mapie

Koszt całkowity
Brak danych

Uczestnicy (6)