Advances in digital technologies and instrumentation have made possible to measure and monitor a wide range of variables in industrial operations and the mining industry is not an exception. However, there are numerous important parameters that still cannot be measured directly due to the measurement's complexity, lack of appropriate access or harsh environment for measurement. An example is the estimation of ball and total filling in Semi-Autogenous Grinding (SAG) mills. Soft sensors can use measurements of process parameters from existing instrumentation along with semi-fundamental mathematical models to calculate a parameter that is not practical to measure directly. This Webinar demonstrates the value of integrating expert knowledge in the form of mathematical models with the data that operating plants generate daily to provide an accurate estimation of key operating parameters as soft sensors.