Soft Sensors
Soft Sensors predict plant quality variables in real-time, allowing for continuous estimation of product properties that are impractical or impossible to measure online.
MACS Soft Sensors
Build Models from Live Data
Soft Sensors read live data from the plant, such as temperatures and pressures, and use them with user-developed models to calculate estimated stream properties. Soft Sensor models are based on first principles or statistical methods such as PCA, PLS and Time Series Analysis using PARCmodel and MACSsuite.
Include Data From Any Source
Simplify data access for model building and online calculations with PARCview. Both PARCmodel and MACSestimator take advantage of the PARCview data architecture to easily read from any available data source, including OPCDA, OPCHDA, plant historian systems (PI, PJD, IP.21, etc.), SQL, Excel and more.
User Friendly
PARCview provides a familiar user interface for model creation and monitoring. Embed your Soft Sensor model in a PARCview trend for model prototyping as well as real-time, open-loop feedback to the operator.
Closed Loop Control
Leverage your Soft Sensor models for use in advanced control applications. Soft Sensor models can be directly exported to MACSestimator, our online model calculation engine for closed-loop control. The Soft Sensor output can be used like any other process value in the control solution.
Advanced Options
Increase the robustness of your online models with dynamic updating based on lab analysis, data validation using PCA and PLS statistical methods, signal filtering and dead-time compensation.