Automated Calibration\nfor Rapid Optical Spectroscopy\nSensor Development for Online Monitoring

An automated platform has been developed to assist researchers\nin the rapid development of optical spectroscopy sensors to quantify\nspecies from spectral data. This platform performs calibration and\nvalidation measurements simultaneously. Real-time, in situ monitoring\nof complex systems through optical spectroscopy has been shown to\nbe a useful tool; however, building calibration models requires development\ntime, which can be a limiting factor in the case of radiological or\notherwise hazardous systems. While calibration time can be reduced\nthrough optimized design of experiments, this study approached the\nchallenge differently through automation. The ATLAS (Automated Transient\nLearning for Applied Sensors) platform used pneumatic control of stock\nsolutions to cycle flow profiles through desired calibration concentrations\nfor multivariate model construction. Additionally, the transients\nbetween desired concentrations based on flow calculations were used\nas validation measurements to understand model predictive capabilities.\nThis automated approach yielded an incredible 76% reduction in model\ndevelopment time and a 60% reduction in sample volume versus estimated\nmanual sample preparation and static measurements. The ATLAS system\nwas demonstrated on two systems: a three-lanthanide system with Pr/Nd/Ho\nrepresenting a use case with significant overlap or interference between\nanalyte signatures and an alternate system containing Pr/Nd/Ni to\ndemonstrate a use case in which broad-band corrosion species signatures\ninterfered with more distinct lanthanide absorbance profiles. Both\nsystems resulted in strong model prediction performance (RMSEP <\n9%). Lastly, ATLAS was demonstrated as a tool to simulate process\nmonitoring scenarios (e.g., column separation) in which models can\nbe further optimized to account for day-to-day changes as necessary\n(e.g., baseline correction). Ultimately, ATLAS offers a vital tool\nto rapidly screen monitoring methods, investigate sensor fusion, and\nexplore more complex systems (i.e., larger numbers of species).