Evaluation of context management architectures: The case of context framework and context broker

In recent years Context Management Systems have been increasingly adopted for processing sensory information in different fields of applications, such as surveillance of construction areas, coordination of emergencies and production monitoring. While system architectures have been designed for specific domains, literature informs only little about evaluation criteria for Context Management-Architectures. Nevertheless it is crucial to use the right criteria, when the system has to face requirements of industrial applications. While one system may be designed for reusable contextualisation mechanisms, others provide real-time information for Manufacturing Execution Systems (MES). In this paper we develop suitable scenarios and metrics for architecture evaluation and apply these on two exemplary real-world architectures. The development of scenarios was based on domain experts and researchers in a project 1 concerning Cyber Physical Production Systems (CPPS), funded by the German Federal Ministry for Economic Affairs and Energy. Our results show, that early architectural decisions have a high influence on the systems performance and therefore the systems applicability. The developed criteria also provide knowledge for system engineers in industrial practice.