Introduction: FormuSense

The ProSensus FormuSense software was developed in order to enable users to explore and develop new product formulations by leveraging available data on existing formulations.

The main features of the software include:

  • Data set assembly:

    • Import product formulations (recipes), ingredient properties, quality data (including replicates), and process data (static and/or time-series.

    • Flexibility for data editing, data value replacement, and expansion of categorical variables.

    • Visualize time-series process data, and calculate key features to include in modeling.

    • Statistics and graphical visualization of formulations, data frequency, and data integrity metrics.

    • Generation of ingredient ratio tables, class ratio tables, and mixture property tables.

    • Options for mixture property configuration including mathematical rules and global property calculations.

    • Graphical visualization and statistical analysis of mixture property results, including data integrity metrics.

    • Generation of data transformations to handle non-linearities and interactions, with options for equations, property selection, and variable ranges.

  • Multivariate modeling:

    • Select variables and formulations to include or exclude in each model.

    • Adjust the number of principal components.

    • Evaluate model fit.

    • Evaluate model predictive ability by specifying a validation group.

    • Evaluate multiple model configurations and locate gaps in data or design space with intuitive graphical visualizations.

  • Simulation:

    • Specify input values (X) for any model to project the results onto the latent space.

    • View resulting model validity metrics, formulation cost, and mixture property values.

    • Predict values for all outcome variables (Y) in a PLS model.

    • Find the existing/historical formulation that most closely matches a simulation configuration.

  • Optimization:

    • Specify quality variable targets (PLS) or score space targets (PCA or PLS) to determine the optimal input values (X) to achieve those targets under given constraints.

    • Impose bound constraints and adjust tuning parameters for the optimization algorithm.

    • Balance or prioritize product quality and raw material costs.

    • Find the existing/historical formulation that most closely matches each optimization solution.

  • Export to Microsoft® Excel® and ability to copy tables at each stage.

Contents: