In a future networked physical world, a myriad of smart sensors and actuators assess and control aspects of their environments and autonomously act in response to it. Examples range in telematics, traffic management, team robotics or home automation to name a few. To a large extent, such systems operate proactively and independently of direct human control driven by the perception of the environment and the ability to organize respective computations dynamically. The challenging characteristics of these applications include sentience and autonomy of components, issues of responsiveness and safety criticality, geographical dispersion, mobility and evolution. A crucial design decision is the choice of the appropriate abstractions and interaction mechanisms. Looking to the basic building blocks of such systems we may find components which comprise mechanical components, hardware and software and a network interface, thus these components have different characteristics compared to pure software components. They are able to spontaneously disseminate information in response to events observed in the physical environment or to events received from other component via the network interface. Larger autonomous components may be composed recursively from these building blocks. The paper describes an architectural framework and a middleware supporting a component-based system and an integrated view on events-based communication comprising the real world events and the events generated in the system. It starts by an outline of the component-based system construction. The generic event architecture GEAR is introduced which describes the event-based interaction between the components via a generic event layer. The generic event layer hides the different communication channels including ∗ This work was partially supported by the EC, through project IST-2000-26031 (CORTEX), and by the FCT,

The paper describes the event model and the architecture of the COSMIC (COoperating SMart devICes) middleware. Based on the assumption of tiny smart sensors and actuators, COSMIC supports a distributed system of cooperating autonomous devices. COSMIC considers quality of service requirements in the event model and provides an application interface which allows to express the respective temporal and reliability attributes on a high,application related abstraction level. According to the need in most realtime systems,COSMIC supports event channels with different timeliness and reliability classes. Hard real-time event channels are considered to meet all temporal requirements under the specified fault assumptions. The resource requirements for this type of channel are statically assigned by an appropriate reservation scheme. Soft real-time event channels are scheduled by their deadlines,but they are not guaranteed under transient overload conditions. Non-real-time event channels are used for events without any specified timeliness requirements in a best-effort manner. The paper finally presents the layered COSMIC architecture to map the different