@ GEPSI –CIEPQPF
The Process Chemometrics Laboratory (PCLab) aims to develop, test, validate and deploy methodologies that support process supervision, diagnosis and improvement, through an effective, thorough and robust use of process data, in whatever structure and format it is available.
Process improvement takes place by alternating deductive (from hypothesis, concepts and models to processes and data) and inductive (from data back to concepts again) stages, until a satisfactory solution is attained that meets the organization goals and process restrictions. Deduction regards the derivation of optimal procedures on the basis of a priori knowledge (assumptions, hypothesis and process models). Induction involves the inference of knowledge from data collected in real processes that will consubstantiate new assumptions, hypothesis and refined models that better describe reality. Process Systems Engineering (PSE) activities require this cyclic pattern, which is, in fact, the underlying mechanics of the scientific method.
However, most of the PSE methods currently proposed are essentially of a deductive nature, and lack the complementary inductive methodologies to close the improvement cycle.
At PCLab, we develop inductive PSE approaches as well as address their proper integration with deductive tools, in order to propose robust and effective solutions to relevant industrial problems.