Over the past few years, the PI System within this paper mill grew from one computer on one machine in the plant, to deploying the PI System in all process areas. The PI System currently captures data from multiple data sources, such as process control systems, online visual inspection systems and quality databases. Spoil remains the most significant cost to the business. As the company progressed towards Operational Excellence, they focused on built-in quality. Their data solutions needs moved away from Process Engineers analyzing trends and reviewing data, toward an enterprise solution where operational managers and crews could recognize and address issues in a standardized way. Now, the mill is undergoing a change in mindset to establish an enterprise culture enabling data-driven decisions. Operators, not just managers, are now asking questions of the data, wanting to expand on the visualizations they already have. This transformation from Information-for-Engineers to Message-to-Operators required a robust and exhaustive approach to data analysis, or a user-friendly approach to operator acceptance. The PI System Asset Framework provided the needed consistency and ease of replication. With Asset Analytics and MKS Umetrics SIMCA, complex calculations and logics were developed. Using analytics to transform raw data into simple messages is at the core of this change. A visualization solution was provided by creating a unique PI ProcessBook display, capturing all the decision trigger points, and allowing further investigation using the classic PI ProcessBook trends displays.
Speaker
Laurent Watremetz
Laurent is responsible for product quality control and process monitoring/improvement activities, whilst also driving and developing data strategy and systems. He has worked on product and process improvement within the Optical Fibres, Food and Beverage and now Pulp and Paper industries. Laurent now focuses on the development of the PI System Data Infrastructure across all on site processes. More recently this has involved the advancement of multivariate tools (SIMCA) in conjunction with the PI System, and the interfacing the PI System with other onsite systems and databases. Laurent holds an M.S.c in Chemistry of Advanced Materials from UMIST, University of Manchester and Superior Diploma of Engineering Material Sciences from ENSI de Caen, France.