Track: Data Analytics

Abstract

 In largely autonomous process control ever-increasing amounts of data are generated and made available for specific analysis functions. This is especially important in the field of sensor-based process control and for predictive analysis of machine data, server logs and text messages. A key challenge is to provide appropriate data processing solutions for effective and efficient integration of data and process management and appropriate analysis tools. Despite considerable advances in Big Data and IOT technology useful solutions are hardly available. A study conducted since 2013 examines the integration capabilities of process management platforms and the most noted Big Data reference platform Hadoop. As a result of this research an easy-to-use Integration library has been developed and released. The library comprises software adapters that allow the seamless integration of Hadoop tools into existing or even new business processes. The work has been validated with the implementation of use cases for the application areas “Predictive maintenance through analysis of operational machine data” and “Failure prediction with analysis of log data in a cellular network”. Further research aims at outlining future applications resulting from the expected symbiosis of these technologies.