This is the authors' abstract. We don't add key points for this paper.
A chemometric-assisted strategy was developed for accurate, online tap water conductivity monitoring using electrical impedance spectroscopy in a low-cost Internet-of-Things device (NEMO). The study focused on optimizing instrumental settings and quantification models to achieve reliable conductivity measurements in the 15–750 μS/cm range, representative of real tap water conditions. Design of Experiments (DoE) was applied to evaluate the influence of key measurement parameters, while Principal Component Analysis (PCA) guided the selection of representative water samples, based on the chemical variability of tap waters in the Lombardia region (Italy). Partial Least Squares (PLS) regression was then used to model the full impedance spectra (2–150 kHz), achieving accurate conductivity predictions directly from raw data. The optimized model, built with data from 144 real water samples and validated on independent test sets, demonstrated excellent analytical performance (RMSECV = 35 μS/cm, RMSEP = 32 μS/cm) and high robustness under simulated distribution system conditions (30 L/h flow rate, 22 °C). Analytical figures of merit, including limit of detection (33–51 μS/cm) and quantification (98–154 μS/cm), were calculated following IUPAC-consistent chemometric procedures. This approach enables high-performance conductivity monitoring using compact and energy-efficient sensors, bridging the gap between device simplicity and analytical accuracy for online water quality monitoring systems.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Bova et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: