Smart vs. Big Data Analysis: How could you get more out of less in board strength modelling?, Smart vs. Big Data Analysis: How could you get more out of less in board strength modelling?Smart vs. Big Data Analysis: How could you get more out of less in board strength modelling?, 19PaperCon

For troubleshooting and optimizing the traditional paper- and board-making processes, the procedure has been to gather as many spread data as possible and find out how to make sense of them regarding possible inputs and respective outputs. After that, some level of filtering is applied in order to recognize what is within average performance range and what is outside it. In that sense, software experts produce plotted reports for paper- and board-making experts to give them feedback and insights about what to do (and where to do it in the process), helping them recognize which variables to tweak and how to proceed. This requires a significant amount of resources (i.e., measurements and instruments), as well as time and multiple iterations until root-cause relationships are found. 

Big Data handling includes the development of soft sensors and a high level of integration at a mill site to provide the opportunity to potentially determine or predict the response in real time when and if the correct set of variables have been preselected. However, it is not easy to determine the correct set of variables for a multi-linear regression model, and it must be remembered that first, second, and third order of magnitude phenomena can appear in combination with strong “background noises” from the process. For this reason, instead of Big Data, we would like to talk about Smart Data. Smart Data allows for modelling, diagnosis, analysis, etc., with an acceptable level of accuracy. 

This article attempts to predict strength responses in paper- and board-making by addressing a minimum set of variables that would have the most relevant impact from process to final product, accounting for first principle criteria and recognizing the second order responses (variability and background noise). It is expected that optimal results might not be achieved in these simulations on the first attempt, but possible alternative tools and methods can be discussed and/or proposed for future work to ultimately provide improved solutions. 

TAPPI conference proceedings and presentations, technical papers, and publication articles provide technical and management data and solutions on topics covering the Pulp, Paper, Tissue, Corrugated Packaging, Flexible Packaging, Nanotechnology and Converting Industries.

Simply select the quantity, add to your cart and your conference paper, presentation or article will be available for immediate download.