The overall hypotheses of the project are:

• Service composition can be supported by generic solutions in the form of methods and
  middleware that significantly reduce the complexity of developing composite services.
• Collaboration-based composition can be applied at runtime and enables the flexible adaptation and composition of collaborative services to users’ goals.
• Ontologies can be exploited to describe users’ goals and to augment collaborative service runtime descriptions with goals, facilitating effective semantic service discovery and automated composition of service collaborations.
• Visual formalism and tools can be developed that support end users in composing service collaborations. These hypotheses will be broken down in a set of sub-hypotheses which will guide the work in a scientifically sound manner.

Validation of research hypotheses through experimentation

The requirements to the composition framework will be extracted from the definition of scenarios and related research work results including the results developed by the partners in previous research projects. The solutions developed in UbiCompForAll will be validated through the realisation of SW prototypes and pilot services. The development of pilot services will provide feedback for the refinements of the requirements to the composition framework and of the research hypothesis.

The project adopts an incremental approach. The project defines five research cycles, where the three intermediate cycles involve the creation of new theoretical results, their implementation and assessment:

Cycle 1 - Initial platform with support for service collaboration modelling, implementation and execution. This cycle aims at building a common basis from the different building blocks provided by the partners.

Cycle 2 - User-driven service composition at design time. This cycle is about providing the end users support for browsing for service components, selecting them, expressing simple service compositions and deploying the software. Support for validating services before deployment will be provided.

Cycle 3 - High-level programming for service composition at design time. This cycle focuses on more advanced support for service composition that requires component configuration and writing "glue" code. Support for service validation will be extended.

Cycle 4 - User-driven service composition at runtime. This cycle focuses on configuration of the composed services after their deployment. Support for service validation at runtime will be provided.

Cycle 5 - Finalization. This cycle is about analysing the results of the project, packaging the final results so they can be more easily taken up by the research community. This cycle will also plan the experimentation phase.