The main DIVA objectives are:

1. To provide novel build time and runtime management of adaptive system (re)configuration of co-existing, co-dependent configurations that can span across several administrative boundaries in a distributed, heterogeneous environment.
2. To provide efficient handling of the number of potential configurations, that may grow exponentially with each new variability dimension,
3. To increase quality and productivity of adaptive system development and help the designers to model, control and validate adaptation policies as well as the trajectory going from one safe configuration to another.
4. To demonstrate its interest and generality and disseminate its results.

Run-time management

The first objective will be accomplished by leveraging the synergy between model-driven and aspect-oriented techniques. There will be a focus on separating the application-specific functionality from the adaptation concerns. Aspect-oriented techniques will be utilised to analyse and reconfigure crosscutting features dynamically. Model driven techniques will be used to raise the level of abstraction and to provide models at runtime. In DiVA, runtime models will expose access points for dynamic manipulation and adaptation. The current AOSD[1] and MDE[2] techniques will be extended with support for dynamic variability and for reasoning about problems of co-dependent co-existing system configurations.

Managing the exponential growth

The second objective will be addressed by efficient analysis, design and runtime representations of potential configurations using aspect-oriented modelling techniques and by using model composition as a variability mechanism. Furthermore, DiVA will provide analysis techniques to reduce the number of configurations to the most pertinent ones.

Increased system development productivity

The third objective will be accomplished by providing an integrated toolset, a methodology and frameworks. The variability and validation analysis will be supported by a methodology for development of variability requirements, design and code artefacts, and analysis of their configurations and co-dependencies, and selection of preferred alternatives, thus, progressively reducing the alternatives’ space. Moreover, DiVA intends to provide techniques for verifying and validating the adaptation both in terms of realising the goals of the adaptation and its trajectory.

Pilot case studies

The fourth objective will be accomplished by implementing case studies from two different domains: crisis management and Customer Relationship Management (CRM), and by disseminating the results through various dissemination channels including publications at international conferences and workshops, contribution of technologies to the open source communities.