The main conceptual blocks and their roles in the realization of the PROSENSE platform, an event-based platform for integrating heterogeneous real-time and dynamic streams created by hardware sensors, software and external data used in manufacturing enterprises. Our conceptual architecture is strongly oriented on the OODA loop and combines smart sensing, anticipation management, incremental proactivity and proactivity management.
Adapters were developed for all necessary enterprise information sources such as hardware sensors (which might include high-precision camera data, accelerometers, vibration and temperature sensors), software sensors from ERP systems and external business context data.
Pub-sub middleware provides the SOA and EDA infrastructure for components and end user services. It is based on an open source ESB for large SOA architectures and fully compatible with all major industry standards (such as JBI, SCA, BPEL or WSDL). If necessary, this message broker will be extended with a novel, dynamic content-based pub/submechanism, required for dynamic CEP.
Smart sensing services enables semantic enrichment with background knowledge and data mining on real-time streams, received through the communication layer. In addition, those services will perform sensor network analytics by using new multi-level algorithms in order to optimize the quality and power of the signal potentially coming from any part of the aggregated space of the raw data.
The Scalable Event Storage provides storage and forwarding of events (in the form of RDF triplets) received from pre-processing service. It is realized as an event cloud, a scalable, semantic, P2P based repository that delivers RDF events to the requesting parties (subscribers) in a push fashion be they external services or project related components, and at
the same time stores the events for historical and statistical purposes. It supports synchronous and asynchronous queries expressed in a subset of the SPARQL language and accessible through corresponding APIs.
Services for Anticipation Management will enable the generation of real-time, data-driven predictions, as well as the discovery of unusual situations, based on events delivered by storage. Novel prediction services will be realized as intelligent services on the top of Probabilistic Stream Processing, a theory basically implemented in Qminer.
Goal-driven Complex Event Processing component has the role of dynamic definition and detection of complex events and reasoning over events, supplied by Event Storage. Complex Event Pattern can be defined dynamically or produced by Anticipatory services (predictions). This research will extend our work on scalable, logic-based event processing, implemented in ETALIS, an open source CEP engine, in the direction of extending it to enable dynamic (onthe-
fly) deployment of new patterns and their monitoring. In addition, this component will provide a methodology for the structured, business goal-driven identification of relevant situations of interest and will leverage detection of anomalies in real-time.
Incremental Proactivity subscribes for interesting complex events (pub/sub communication) and generates corresponding recommendations, by taking into account business context. It couples typical approaches in recommender systems with decision making methods.
Proactivity Management deals with the KPI modeling languages in order to leverage continuous business improvement based on the recommendations.