Did you think that to master Multi-Variable Statistical Process Control (MSPC) you would have to be an expert in multivariate techniques including Partial Least Squares, Principal Components, Neural Networks, Recursive Partitioning, Independent Components Analysis, Cluster Analysis and other forms of statistical analysis?

Curvaceous Software has brought the incredible power of multivariate modelling and analysis out of the domain of master statisticians and onto the engineer's desktop.

This wholly new technology (Geometric Process Control - GPC) requires no maths expertise and no statistical analysis as it is entirely visual. Unlike previous Multi-Variable Statistical Process Control methods hundreds of variables and their interactions can now be seen on one interactive display. Non-linear models of complex batch or continuous processes can be auto generated from the data-set. These models can then be used to guide operations to always keep the product quality in spec.

Multi-Variable Statistical Process Control (MSPC) has never been so easy, as over 80 process companies have already discovered.

So stop struggling with dimensionality reduction, principal component analysis, multi-variable graphs, obscure and complicated neural-nets and other forms of complex statistical analysis and contact us to demonstrate how a non-linear process with 100 variables can be modelled in minutes.

To the right is a simple batch process, that was modelled in this way. It has been speeded up X1000. The dots are the real-time process values on vertical scales. The green lines are the dynamic control limits. If the dots move out of the green envelope, the process goes out of capability. Watch how the useable ranges of the variables change throughout the batch.

This model was made in under an hour from existing plant data. How long would it take with traditional methods?
 

If you want to see Geometric Process Control technology in action with your data or have any questions please contact us.