Self-adaptive software is a very active field of research. Many projects (also sponsored by the European Commission (EC)) have been funded with the goal of providing techniques and tools for self-adaptive (and autonomous) programs. This paper is about tools for the development and maintenance of self-adaptive software. It specifically addresses tools using adaptation engines, partially described using the feedback loops approach.

The paper exhaustively presents the details of the executable runtime megamodels (EUREMA) tool. EUREMA is a “model driven engineering approach ... that enables the specification and execution of adaptation engines for self-adaptive software with multiple feedback loops.” An interesting feature of EUREMA is a language that can be used to specify the aforementioned feedback loops. It is used to define how the models and the software relate to each other at runtime, and when the context changes and the software needs adaptation.

Of course, adaptation cannot be solely provisioned at runtime. If the context changes heavily, or if the context alters itself to an unexpected or unplanned situation in the loop, developers must intervene and provide maintenance support. This kind of offline adaptation is also covered by the EUREMA tool, which enables its users to use the same model for both offline and online adaptation.

The overall toolset seems to be extremely promising in the field of self-adaptation and offline adaptation. The syntax is intuitive as it is highly inspired by the unified modeling language (UML). EUREMA has been implemented using the Eclipse modeling framework (EMF) in Java, and it is available as an Eclipse plugin from an Eclipse update site. The toolchain is under active development.

The authors also carried out a set of performance evaluation tests, with good results. They prove that the overhead introduced by the EUREMA tool during self-adaptation is low.