This application development tool can be used to target the SpectraWiz® process monitor/control environment or can be customized for specific OEM applications.  Neural Nets require development of an analytical setup that provide a quantified spectral dataset for the target sample(s).  The Neural Network can then predict the concentration of unknowns such as % alcohol in a process stream.  Neural net development uses quantified spectral data to "train" itself, resulting in accurate predictions of unknowns within desired tolerance requirements.  The concentration relationships with respect to spectral absorbance, transmittance, reflectance, emittance or fluorescence, CAN BE non-linear.

StellarNet utilizes the California Scientific Software's BrainMaker package to perform the Neural Network training and development.  Our SpectraWiz(tm) software has the ability to execute the Neural Network(s) developed for specific process and OEM applications in real-time.   StellarNet offers all components required as a complete development resource.