Multivariate Data Analysis Software Models Batch and Continuous Processes

Umetrics released the latest version of its SIMCA multivariate data analysis software for batch and continuous process modeling. The software can use input from spectral data (e.g., Raman spectroscopy) as well as conventional process sensors (e.g., temperature, pressure) in models of unit operations and for complete processes. New features such as scripting, which allow users to automate routine procedures, are designed to save time for users when building models. Scripts can be shared across an organization to streamline data analysis tasks. SIMCA 14 also has improved notes and reporting capabilities and a "Workspace" feature that saves a user's customized plots and tables. "What if" analysis use of model to predict what will happen when data is adjusted by simulating changes to parameters, such as a temperature or flowrate.

Multivariate predictions, using models built in SIMCA, enable real-time monitoring of processes using SIMCA-online. This system produces control charts that summarize the state of the process. Analysis of deviations and alarms are provided through interactive drill-down features and are available for both historical and forecasted observations. The Control Advisor in SIMCA-online 13.3 includes a Forecast mode for predictive monitoring and an Advised Future mode, which can be used to optimize a process with model predictive control by suggesting process adjustments that may be implemented manually or automatically.