Software sensor for potable water quality through qualitative and quantitative analysis using artificial intelligence

The analysis and control of potable water quality is increasingly fascinating due to its impacts on human life. Numerous lab-scale and field-scale treatment and sensing methods are created in this field to safeguard this natural vital asset. From long several methods were experimented determining water quality including traditional one´s such as wet-chemistry which needs reagents, electro-chemical based, and most recently machine learning based software models to name a few however, performance enhancement and development of truly ion-selective electrodes has been still area of most interest and current area of research world-wide. In this paper, spectroscopic fusion for quantitative determination of qualitative attributes of water parameters will be explored with the application of chemometrics. An integration of multi-spectral, surface enhanced Raman spectroscopy, UV-Visible spectroscopy in the presence of multi-sample holder made off with and without nanostructured substrate will be attempted, and the patterns would be analyzed using Principal Component Analysis and other similar Machine Learning techniques. A set of pseudo-sampling matrix comprising of training and validation sets would be demonstrated on a lab-scale basis as a proof-of-concept. This paper also aims to overview existing practices, and presents proposed approach which would be free from reagent, rugged, and field-usable method, and would use fusion of spectroscopy, nano-structured sample holder, and Machine learning extraction algorithms.