Role-Based Models for Building Adaptable Collaborative Smart Service Systems

The ability of service systems to compose their structure on-demand at run time and to change their application behavior in response to changes in their computational environment will become increasingly important. Approaches related to self-organizing software systems focus on automated composition of dependent service systems at run time but lack the specification of complex service structures whereas context-dependent system changes are usually limited to adaptations local to a subsystem and lack the ability to perform complex modifications in a decentralized infrastructure. Hence, in this paper, we bridge the gap between the specification of complex, context-aware service structures at design time and their reliable adaptation at run time. We propose a context-aware, role-based collaboration specification to model adaptable, collaborative smart service systems as well as an integrated middleware architecture for on-demand composition and subsequent reliable adaptation of such. We validate our approach using a case-study from the tech- enhanced teaching domain.