Search

The so-coined fourth paradigm in science has reached the sensing area, with the use of machine learning (ML) toward data-driven improvements in sensitivity, reproducibility, and accuracy, along with the determination of multiple targets from a single measurement using multi-output regression models. Particularly, the use of supervised ML models trained on large data sets produced by electrical and electrochemical bio/sensors has emerged as an impacting trend in the literature by allowing accurate analyses even in the presence of usual issues such as electrode fouling, poor signal-to-noise ratio, chemical interferences, and matrix effects. In this trend article, apart from an outlook for the coming years, we present examples from the literature that demonstrate how helpful ML algorithms can be for dispensing the adoption of experimental methods to address the aforesaid interfering issues, ultimately contributing to translate testing technologies into on-site, practical, and daily applications., (© 2023. Springer-Verlag GmbH Germany, part of Springer Nature.)

Flow assurance plays an important role in designing safe and efficient operation techniques along oil and gas explorations. In this way, approaches able to provide fast and field-deployable monitoring of scaling ions in crude oils may benefit the petrochemical industry in its long-standing mission to devise ideal flow conditions by aiding the adoption of optimum dosages of off-the-shelf products. We first describe a scalable platform toward rapid (∼10 min) determination of multiple scaling ions in crude oils. Our platform is based on turbulence-mediated ultrafast and efficient microfluidic solvent extraction (μSE) combined with impedimetric sensors and machine learning to determine different ions from a single impedance plot. 3D-printed μSE devices were able to afford ultra-fast (residence time of ∼1.0 s) extractions. Nanocellulose-based foams allowed us for rapidly separating the water-oil phases, with the aqueous phase being sampled for posterior detection. Scalable sensors obtained by distinct prototyping methods provided the multidetermination of ions in 50 produced water samples (Mg2+, Ca2+, Sr2+, and Cl⁻) and 49 crude oils (Ca2+ and Cl⁻). The accuracies ranged from ∼96 to 100%. Importantly, ML models trained on standard solutions delivered poor accuracy, showing the relevance of learning supervised algorithms with real samples to deliver accurate capacitive analyses. • A simple platform granted the determination of multiple scaling ions in crude oils. • Ultrafast and high-efficiency microfluidic extraction of ions from oil is achieved. • Nanocellulose foams allow fast separation of W-O phases for following water analysis. • Multiple ions are determined by combining impedimetric sensors with machine learning. • All the experiment can be completed within 10 min [ABSTRACT FROM AUTHOR]