Process Chemometrics

Process Chemometrics (or Process Analytics) is a highly interdisciplinary field which embodies methods and approaches from Process System Engineering, Multi- and Megavariate Statistics, Mathematics, and Computer Science to address relevant problems in industrial processes or complex systems. Even though its roots can be traced back to the late 1950’s, the area begins to expand in scope and relevance in the early 1990’s. Today, Process Chemometrics provides a body of knowledge capable to deal with the main features of industrial data in activities such as process {modelling, monitoring, control, optimization and design}, product design and development, transfer of knowledge between related industrial sites, troubleshooting and diagnosis, among others.

Some of the current challenges in this field include:
• Optimal fusion of data with different structures (sensors, spectra, images,…);
• Development of better diagnostic methodologies capable to overcome some of the limitations of current methods, such as the smearing problems of contributions plots;   
• Combination of a priori and data-driven knowledge (hybrid methods, grey models, Bayesian approaches,… );
• Integration of statistical and automatic (or engineering) process control (SPC-EPC);
• Explicit Incorporation of uncertainty information in process chemometric methods;
• Use of local and global (or marginal) measures of association to describe the system structure.