As the number of elderly people within the EU are significantly growing the EU faces a major challenge in the upcoming decades to support their well-being. However, it requires a lot of attention and time from doctors to provide feasible treatments and monitoring for these people that requires novel approaches to keep up with their growing number. Telecare systems offer solution for these problems, as they provide advanced monitoring capabilities from simple daily blood-pressure measurements to long-term blood-pressure characteristic based prediction that might trigger pre-emptive intervention.
The aim of the European Union funded "CONCERTO" project is to realize a model-driven framework, based on the CHESS methodology, for developing and validating safety-critical embedded system using a common set of modeling based techniques that can be reused in the different safety-critical application domains (e.g., avionics, petroleum, automotive, space, etc). Within the project the role of BME was to develop a complete code generation tool-chain for telecare systems.
The current thesis describes the model translation steps developed for the CONCERTO based telecare tool-chain. The approach follows the CONCERTO methodology that (i) uses a subset of UML component diagrams to define the platform independent descriptions of the systems then (ii) in separately defines the underlying execution platform using also UML component diagrams, finally (iii) the platform independent models is mapped to the execution platform from this integrated model the platform specific representation is generated, which serves as the input for the dedicated code generators. As the telecare domain required specific modeling concepts such as triggers and reactive execution mechanisms a telecare and platform specific model was defined to capture these aspects. For demonstration purposes the translation itself was developed using two different approaches one is the conventional Java based approach where all transformation steps are defined in pure Java, while the second approach is based on the EMF-IncQuery declarative query technology that provides incremental query capabilities over graph models. Finally, these two approaches are compared both in usability aspects and runtime characteristics to help future CONCERTO developers when selecting transformation technology.