Early detection and decision support for critical situations in the production environment

Persons in charge

Hassan Enam Al Mawla, M. Sc.

Alexander Rehmer, M. Sc.

David Arengas, M. Eng.

Benjamin Jäschke, M. Sc.

Duration

September 2014 - December 2017

Sponsor

Federal Ministry of Education and Research

Brief description

In many industrial sectors, a high degree of automation enables an efficient operation even in countries with high wages such as Germany. This has led to plant operators having less opportunity to experience the dynamic behavior and operation of industrial processes. As a result, they can lack the necessary experience to cope with unexpected operating conditions and become overwhelmed by the sheer number of alarms and malfunction messages. This loss of control greatly increases risk to humans and the environment, may cause damage to valuable assets and can result in production losses and plant downtime.

The FEE joint project addresses the goal of bringing together heterogeneous data from sensors, engineering databases, process information management systems, shift logs and operating instructions recorded at production sites over a number of years. This information will be entered into a joint analysis platform with the objective of assessing the possibility for automated integration. Based on this platform and with the help of big data approaches, real time methods will be developed to warn operators of potential problems at an early stage and provide assistance functions to support them in the development of intervention strategies. In this manner, it should be possible to move from reactive to proactive intervention.

The project consortium consists of research, industrial development and application partners. The consortial members collectively cover all the necessary scientific and industrial skills in the areas of big data computation, data mining, process automation systems, system dynamics, human-machine interaction as well as plant maintenance and operation.

The Department of Measurement and Control at the University of Kassel is primarily concerned with the detection and dynamical modeling aspects of the project, including:

  • Developmentof methodsfor the earlydetectionof critical situations for big data application through the use of time series analysis methods
  • Development of methods for operator decisionsupport in critical situations for big data application by meansofsystem identificationmethods
  • Test and demonstration of the developed methods for the model factories of the University of Kassel and Dresden University of Technology as well as for historical data from production plants of the associate partners

Cooperation partners

  • ABB AG
  • Chair of Knowledge and Data Engineering, University of Kassel
  • Chair of Process Control Systems Engineering, Dresden University of Technology
  • RapidMiner GmbH

Associate partners

  • PCK Raffinerie GmbH
  • INEOS Köln GmbH
  • BASF SE

Student contributions

Please click on the following link for student contributions and refer to projects and theses supervised by the persons in charge of this project:

http://www.uni-kassel.de/maschinenbau/en/institute/mess-und-regelungstechnik/student-works.html

External links