TU München - Fakultät für Informatik
Software- and Systems Engineering Research Group

  Home  |  Forschung/Kompetenz  |  Lehre  |  Personen  |  Publikationen  |  Abschlussarbeiten  |  Sonstiges  |  Kontakt  | 

Agenda

Es sprechen Studenten über ihre abgeschlossenen Diplomarbeiten und Systementwicklungsprojekte.

Am Dienstag, 06.02.18, ab 10:30 Uhr, im Raum „Tarski" (01.09.011B) :

Enabling Stream Analytics in IoT Mashup Tools

Internet of Things (IoT) developers need to integrate various protocols, back-end components, services, and different APIs as well as to pre- and post-process data to build IoT applications. As a result, IoT mashup tools have emerged to provide a flexible, but easy to use visual programming environment for rapid development. However, common IoT mashup tools such as Node-Red do not provide mechanisms to deal with continuous and big amounts of data. Today, the speed at which data are generated, consumed, processed and analyzed is constantly increasing at an unbelievably rapid pace. Therefore, traditional data processing approaches like batch analytics are no longer suited for these data. Consequently, multiple tools have been introduced such as Apache Flink and Spark Streaming that deal with data as streams and offer stream analytics. Yet, these tools can be difficult to use and orchestrate especially for a nonprogrammer. This work tries to solve both problems by enhancing an IoT mashup tool with stream analytics capabilities. In particular, we implement and integrate a stream analytics suite, composed of various pre-programmed stream processing functions, into the aFlux mashup tool, which was developed at the Technical University of Munich. This way, we deliver an easy-to-use tool that allows even non-expert users with no coding skills to analyze streams of data in real-time using a graphical user-interface. On top of that, we offer a number of adjustable options for stream analytics such as overflow strategies and windowing methods. In the frame of this work, we evaluate this tool through two experiments by exposing it to real data streams. The first experiment is about the sentiment analysis of a Twitter feed in real-time whereas the second refers to the data analysis of a traffic simulation. The results show that our tool is fully functional and able to analyze streams of data in real-time.