Runtime Anomaly Detection Method in Smart Factories using Machine Learning on RDF Event Streams: Grand Challenge

This year's ACM DEBS Grand Challenge problem is about anomaly detection of manufacturing equipments based on machine learning techniques, which is fairy challenging. This requires semi-realtime handling of RDF data values continuously collected in streams, measured from analog sensors attached to multiple machines. This paper shows our experience in implementing solutions of the problems in this domain. It includes our elaboration on high degree of concurrency in continuous query processing, to make better use of distributed environments provided by docker containers.