Abstract
Analytical grade 'in vivo' plant metabolic quantification using spectroscopy is a key enabling technology for precision agriculture.Current methods such as PLS, ANN and LS-SVM are non-optimal for resolving spectral interference and matrix effects to provide similar results to the analytical chemistry laboratory. This research presents a new self-learning artificial intelligence (SL-AI) method based on the search of covariance modes. These isolate the different modes of interference present in spectral data, allowing the consistent quantification of constituents. A review of the state-of-the-art methods with the figures of merit mean absolute standard error percentage (MASEP) and Pearson correlation coefficient (R) is presented for comparison and discussion. 707 grapes were quantified for glucose, fructose, malic and tartaric acids in five wine-making and one table grape varieties, and used to benchmark the new method against the state-of-the-art methodologies: partial least squares, local partial least squares, artificial neural networks and least squares support vector machines. SL-AI provides consistent quantifications, whereas previous methods exhibit data-driven performance dependence. Pearson correlations of 0.93 to 0.99 and MASEP of 3.70% to 7.33% were obtained with the new methodology. Local partial least squares, the method with the best benchmarks from literature, achieved correlations of 0.81 to 0.94 and MASEP of 8.00% to 13.4%. The covariance mode isolates a particular interference, providing a direct relationship between spectral inference and constituent concentrations, consistent with the Beer-Lambert law. Such quantifies non-dominant absorbance constituents (e.g. sugars and acids), which is a significant step towards 'in vivo' plant physiology-based precision agriculture.

Abstract
In hydroponic cultivation, monitoring and quantification of nutrients is of paramount importance. Precision agriculture has an urgent need for measuring fertilization and plant nutrient uptake. Reliable, robust and accurate sensors for measuring nitrogen (N), phosphorus (P) and potassium (K) are regarded as critical in this process. It is vital to understand nutrients’ interference; thusly, a Hoagland fertilizer solution-based orthogonal experimental design was deployed. Concentration ranges were varied in a target analyte-independent style, as follows: [N] = [103.17–554.85] ppm; [P] = [15.06–515.35] ppm; [K] = [113.78–516.45] ppm, by dilution from individual stock solutions. Quantitative results for N and K, and qualitative results for P were obtained.

Abstract
This work addresses the use of hyperspectral imaging systems for remote detection of marine litter concentrations in oceanic environments. The work consisted on mounting an off-the-shelf hyperspectral imaging system (400-2500 nm) in two aerial platforms: manned and unmanned, and performing data acquisition to develop AI methods capable of detecting marine litter concentrations at the water surface. We performed the campaigns at Porto Pim Bay, Fail Island, Azores, resorting to artificial targets built using marine litter samples. During this work, we also developed a Convolutional Neural Network (CNN-3D), using spatial and spectral information to evaluate deep learning methods to detect marine litter in an automated manner. Results show over 84% overall accuracy (OA) in the detection and classification of the different types of marine litter samples present in the artificial targets.

Abstract
Plasmonic bio/chemical sensing based on optical fibers combined with molecularly imprinted nanoparticles (nanoMIPs), which are polymeric receptors prepared by a template-assisted synthesis, has been demonstrated as a powerful method to attain ultra-low detection limits, particularly when exploiting soft nanoMIPs, which are known to deform upon analyte binding. This work presents the development of a surface plasmon resonance (SPR) sensor in silica light-diffusing fibers (LDFs) functionalized with a specific nanoMIP receptor, entailed for the recognition of the protein human serum transferrin (HTR). Despite their great versatility, to date only SPR-LFDs functionalized with antibodies have been reported. Here, the innovative combination of an SPR-LFD platform and nanoMIPs led to the development of a sensor with an ultra-low limit of detection (LOD), equal to about 4 fM, and selective for its target analyte HTR. It is worth noting that the SPR-LDF-nanoMIP sensor was mounted within a specially designed 3D-printed holder yielding a measurement cell suitable for a rapid and reliable setup, and easy for the scaling up of the measurements. Moreover, the fabrication process to realize the SPR platform is minimal, requiring only a metal deposition step.

Abstract
Laser-induced breakdown spectroscopy (LIBS) is a technique that leverages atomic emission towards element identification and quantification. While the potential of the technology is vast, it still struggles with obstacles such as the variability of the technique. In recent years, several methods have exploited modifications to the standard implementation to work around this problem, mostly focused on the laser side to increase the signal-to-noise ratio of the emission. In this paper, we explore the effect of pulse duration on the detected signal intensity using a tunable LIBS system that consists of a versatile fiber laser, capable of emitting square-shaped pulses with a duration ranging from 10 to 100 ns. Our results show that, by tuning the duration of the pulse, it is possible to increase the signal-to-noise ratio of relevant elemental emission lines, an effect that we relate with the computed plasma temperature and associated density for the ion species. Despite the limitations of the work due to the low-resolution and small range of the spectrometer used, the preliminary results pave an interesting path towards the design of controllable LIBS systems that can be tailored to increase the signal-to-noise ratio and thus be useful for the deployment of more sensitive instruments both for qualitative and quantitative purposes. © 2022, Universidade do Porto - Faculdade de Engenharia. All rights reserved.

Abstract
One of the caveats of laser-induced breakdown spectroscopy technique is the performance for quantification purposes, in particular when the matrix of the sample is complex or the problem spans over a wide range of concentrations. These two questions are key issues for geology applications including ore grading in mining operations and typically lead to sub-optimal results. In this work, we present the implementation of a class of clustered regression calibration algorithms, that previously search the sample space looking for similar samples before employing a linear calibration model that is trained for that cluster. For a case study involving lithium quantification in three distinct exploration drills, the obtained results demonstrate that building local models can improve the performance of standard linear models in particular in the lower concentration region. Furthermore, we show that the models generalize well for unseen data of exploration drills on distinct rock veins, which can motivate not only further research on this class of methods but also technological applications for similar mining environments.