Former Projects

Jadex Event Stream Processing Architecture

Abstract

The Jadex Event Stream Processing Architecture (JESPA) is a middleware specifically designed to detect and process complex events within streams of primitive events. Such primitive events may originate from RFID systems or any other kind of physical or virtual sensor or sensor network, and usually contain raw sensor measurements, such as an EPC read by an RFID reader.

In order to detect complex relationships between multiple primitive events, possibly spanning different event streams, event stream processing (ESP) and complex event processing (CEP) techniques are applied. For this purpose, the JESPA middleware incorporates a high-performance, open source event processing engine (Esper), which allows to define complex event patterns using an SQL-like event-processing language.

Moreover, the middleware allows to individually pre-process the incoming events before they are forwarded to the complex event processing engine, and to postprocess the resulting generated complex events. For this purpose, a multitude of standard services (e.g., classification, persistence, augmentation) are offered by the middleware, and may be complemented by a producer or consumer by injecting additional user-defined services.

For the orchestration of services within the middleware, executable BPMN-based processing workflows are used. An example of such postprocessing is the escalation handling, during which arbitrary service agents may be invoked in order to react to undesired conditions. The post-processing stage may also be used to create individually tailored application-level events (so called context-based views), which only contain relevant information in a desired format, by invoking conversion, translation, and encoding services. The results can then directly be further processed by enterprise systems, such as supply-chain management and workflow-management systems, intelligent objects, or web-based services. This way, event producers (RFID/sensor readers) and event consumers (like users or enterprise systems) do not need to adapt their business logic to the middleware's functionality and data formats, but may abstract from how events are generated and how actions are carried out, benefiting from the middleware's ability to conduct the overall processing.

Further Information can be found on the JESPA Homepage

Publications of Project

2012		Dirk Bade, Winfried Lamersdorf Integration von Kontextinformationen in Smart Applications und Smart Workflows In: Smart Mobile Apps, Kapitel 20
		Dirk Bade, Daniel Gleim Towards Sensor-supported Indoor Localization Using Cloud-based Machine Learning Techniques In: Tagungsband zum 9. GI/KuVS-Fachgespräch Ortsbezogene Anwendungen und Dienste
2011		Dirk Bade, Roman Puszies A Webservice-based Context Data Service for the Android Platform In: Tagungsband zum 7. GI/KuVS-Fachgespräch
		Steffen Kunz, Benjamin Fabian, Holger Ziekow, Dirk Bade From Smart Objects to Smarter Workflows An Architectural Approach In: Enterprise Distributed Object Computing Conference Workshops (EDOCW), Helsinki
2010		Dirk Bade, Winfried Lamersdorf An Agent-based Event Processing Middleware for Sensor Networks and RFID Systems In: Computer Journal, Special Issue on Agent Technologies for Sensor Networks
2009		Dirk Bade Verteilte Abfragebearbeitung von Sensordaten In: 8. GI/ITG KuVS Fachgespräch Sensornetze, Hamburg, Germany
		Peter Ibach, Dirk Bade, Steffen Kunz Smart Items in Ereignisgesteuerten Prozessketten In: Verwaltung, Analyse und Bereitstellung kontextbasierter Informationen, Workshop, 39. GI-Jahrestagung, 2009, Lübeck, Germany
		Dirk Bade Towards an Extensible Agent-based Middleware for Sensor Networks and RFID Systems In: Third International Workshop on Agent Technology for Sensor Networks (ATSN-09), Budapest, Hungary

11. September 2022 at 14:26 by Prof. Dr. -Ing Norbert Ritter

2013		Diploma Thesis of Nicolas Haase Entwicklung eines Werkzeugs zur Unterstützung von Smart Home-Simulationen Tutor: Dirk Bade Supervisors: Winfried Lamersdorf, Bernd Page
2012		Master Thesis of Cristina Murgu Evaluation von Verfahren zur Datenreduktion für kontextsensitive Systeme Tutor: Dirk Bade Supervisors: Winfried Lamersdorf, Jianwei Zhang
		Bachelorarbeit of Daniel Gleim Sensorgestützte WLAN-Positionsbestimmung durch maschinelle Lernverfahren zur effizienten Indoor Navigation Tutor: Dirk Bade Supervisors: Winfried Lamersdorf, Hannes Federrath
		Bachelorarbeit of Julian Kalinowski Realisierung der Ad-hoc-Kommunikation zwischen aktiven Komponenten auf mobilen Geräten Tutors: Dirk Bade, Alexander Pokahr Supervisor: Winfried Lamersdorf
		Bachelorarbeit of Waldemar Spitschak Delegation datenintensiver Aufgaben von Mobilgeräten in die Infrastruktur Tutor: Dirk Bade Supervisors: Winfried Lamersdorf, Tilo Böhmann
2011		Master Thesis of Richard Günther Integration ereignisbasierter Middleware-Systeme in kontextbasierte Geschäftsprozesse Tutor: Dirk Bade Supervisors: Winfried Lamersdorf, Tilo Böhmann
		Bachelorarbeit of Pascal Folleher Verteilte Klassifizierung von Sensordaten: Ein agentenbasierter Ansatz Tutor: Dirk Bade Supervisors: Winfried Lamersdorf, Jianwei Zhang
		Diploma Thesis of Hai-Minh Le Visuelle Unterstützung bei der EPL-Anfragestellung für Sensordatenströme Tutor: Dirk Bade Supervisors: Winfried Lamersdorf, Norbert Ritter
2010		Bachelorarbeit of Roman Puszies Entwicklung eines Kontextdaten-Frameworks für Android-Endgeräte Tutor: Dirk Bade Supervisor: Winfried Lamersdorf
		Bachelorarbeit of Christoph Koch Migration von Softwareagenten in Jadex Tutor: Dirk Bade Supervisors: Winfried Lamersdorf, Daniel Moldt
2009		Bachelorarbeit of Alex Schajew Entwurf eines Szenario-Simulators zur Unterstützung der Entwicklung von RFID-Anwendungen Tutor: Dirk Bade Supervisors: Winfried Lamersdorf, Klaus Dieter Heidtmann
		Bacc. Thesis of Jan Lueke Filterung von Positionsdaten auf mobilen Geräten unter Verwendung des Douglas-Peucker-Ramer-Verfahrens Tutor: Dirk Bade Supervisor: Winfried Lamersdorf