Project ESMOCYP

What we do

Modern distributed applications where embedded components and networks are used to monitor and control physical processes in the real world (Cyber-Physical Systems - CPS) are usually required to provide semantically enriched yet real-time analysis of sensor streams even in the presence of failures. Theses applications, on the one hand, demand a comprehensive semantic model for real-time data stream analysis, reasoning and actuation on the environment, and correctness of the data stream analysis. To this end, two major problems have to de addressed. Firstly, because of the intrinsic complexity of most CPS (e. g. in Intelligent Transportation Systems) the associated ontology is also very large and there is not yet a scalable semantic model and reasoning technique capable of deducing new knowledge in real-time, which in turn is required for timely reaction to real-world events. Secondly, ensuring correct real-time analysis in the presence of simultaneous failures without compromising the scalability of the system - by inducing high runtime overhead during failure free execution - is still very challenging especially for semantically enriched analysis of sensor stream. Tackling these problems, this project proposes models, mechanisms and algorithms for fault-tolerant data stream processing and real-time reasoning based on the concepts of Semantic Informatin Streams, as a natural extensions of Complex Event Processing (CEP) and RDF (a graph-based knowledge model). The mais advantages of our approach are that time plays a key role in establishing the relation between pieces of information, and that semantic streams can be derived in a two-stage process from the basic events (probled raw sensor data) and consultation of a knowledge base. This knowledge base can also be update with the newly derived facts.

• Learn more