LTH-image

Margret Bauer, University of Witwatersrand

 

Dark data? The use of sensor data in process monitoring and control

Dark Data is a term for underused information that is collected, processed and stored and usually incurs more expense than value. Industrial manufacturing plants are prime suspects for gathering dark data. Many different sensors capture measurements, which is primarily stored in the plant information management system (PIMS). However, analysing the stored data systematically is a difficult task. Examples of making use of dark data are control loop performance monitoring, root cause analysis, sensor validation, predictive maintenance, alarm management and energy management. Increased processing power and big data analytics are opening new opportunities but rely on organised data capture. This talk discusses what methods and technology are currently available to get value out of dark data and shows the limitations of these approaches. It builds on the presenter’s experience at ABB Corporate Research and presents current software solutions and developments.

Presentation slides

 

Biography

Margret is Associate Professor in the School of Electrical and Information Engineering at the University of the Witwatersrand, Johannesburg, South Africa, since 2014.
 
Prior to her current engagement Margret worked for ABB Corporate Research, Germany, focusing on operational technology topics in chemicals, petrochemicals, pulp & paper, minerals and metals since 2005.
 
She has worked and published in the areas of process control, integration of industrial software systems and operations management.
 
Margret received her PhD from University College London, UK, in 2002 and her Master’s degree in electrical engineering from the University of Erlangen-Nuremberg, Germany. From 2005-2006 She was a post-doctoral fellow and recipient of the Claude-Leon-Fellowship as well as the fellowship of the National Research Foundation at the University of Pretoria, South Africa and was appointed as associate professor at the University of Johannesburg, South Africa.
 
Margret’s research interest is on data analytics and big data for process control, recently focusing on the sensor’s lifecycle. She is currently involved in organizing a control conference for Africa and working with local companies on data mining issues.
 
Professional homepage