2018 - PI World - San Francisco - Forest & Paper Products
Feeding the Machine Learning Monster
International Paper’s “Mill of the Future” will change the way we consume process information. One component of our “Mill of the Future” is our advanced data analytics work where we will utilize process and final quality data to better understand and optimize our processes from digesting wood through the final paper production. The original machine learning request was for up to 10,000 process input points for 3-5 years. These decision support and machine learning tools outstripped our existing extraction and analysis techniques which forced a methodology change from PI Datalink spreadsheets to OLEDB extractions to business intelligence tools.
Rick Smith is a Chemical Engineer with 34 years of paper industry experience. His career has always been focused around process control, process data and mill information systems. His focus has long been to get the data to the decision makers to improve product quality and company profitability. Rick has worked with the PI system since the ‘90’s and spend a lot of time in the past five years educating new hires on methods for retrieving process data to solve day-to-day problems.