The great potential of the convolutional neural networks (CNNs) provides novel and alternative ways to monitor important parameters with high accuracy. In this study, we developed a soft sensor model for dynamic processes based on a CNN for the measurement of suspended solids and turbidity from a single image of the liquid sample to be measured by using a commercial smartphone camera (Android or IOS system) and light-emitting diode (LED) illumination. For this, an image dataset of liquid samples illuminated with white, red, green, and blue LED light was taken and used to train a CNN and fit a multiple linear regression (MLR) by using different color lighting, we evaluated which color gives more accurate information about the concentration of suspended particles in the sample. We implemented a pre-trained AlexNet model, and an MLR to estimate total suspended solids (TSS), and turbidity values in liquid samples based on suspended particles. The proposed technique obtained high goodness of fit ($R^2$ = 0.99). The best performance was achieved using white light, with an accuracy of 98.24% and 97.20% for TSS and turbidity, respectively, with an operational range of 0–800 ${\mathrm{mgL}}^{-1}$, and 0–306 NTU. This system was designed for aquaculture environments and tested with both commercial fish feed and paprika. This motivates further research with different aquatic environments such as river water, domestic and industrial wastewater, and potable water, among others.

Light stimulation and biofloc technology can be combined to improve the efficiency and sustainability of tilapia production. A 73-day pilot experiment was conducted to investigate the effect of colored light on growth rates and nutritional composition of the Nile tilapia fingerlings (Oreochromis niloticus) in biofloc systems. The effect of colored light on the nutritional composition of bioflocs as a food source for fish was measured. Three groups were illuminated in addition to natural sunlight with colored light using RGB light emitting diodes (LEDs) with peak wavelengths (λ) of 627.27 nm for red (R), 513.33 nm for green (G), and 451.67 nm for blue (B) light. LED light intensity was constant (0.832 $\mathrm{mW}/{\mathrm{cm}}^2$ ), and had an 18-h photoperiod of light per day throughout the study. The control group was illuminated only with natural sunlight (natural). Tilapia had an average initial weight of 0.242 g. There was a significant effect of colored light on tilapia growth and composition. The R group showed the best growth rate, highest survival, and highest lipid content. The B group showed homogeneous growth with the lowest growth rate and lipid content, but the highest protein level. On the other hand, the biofloc composition was influenced by the green light in the highest content of lipids, protein, and nitrogen-free extract.

Recent research has demonstrated the effectiveness of convolutional neural networks (CNN) in assessing the health status of bee colonies by classifying acoustic patterns. However, developing a monitoring system using CNNs compared to conventional machine learning models can result in higher computation costs, greater energy demand, and longer inference times. This study examines the potential of CNN architectures in developing a monitoring system based on constrained hardware. The experimentation involved testing ten CNN architectures from the PyTorch and Torchvision libraries on single-board computers: an Nvidia Jetson Nano (NJN), a Raspberry Pi 5 (RPi5), and an Orange Pi 5 (OPi5). The CNN architectures were trained using four datasets containing spectrograms of acoustic samples of different durations (30, 10, 5, or 1 s) to analyze their impact on performance. The hyperparameter search was conducted using the Optuna framework, and the CNN models were validated using k-fold cross-validation. The inference time and power consumption were measured to compare the performance of the CNN models and the SBCs. The aim is to provide a basis for developing a monitoring system for precision applications in apiculture based on constrained devices and CNNs.

The automotive industry is entering a digital revolution, driven by the need to develop new products in less time that are high-quality and environmentally friendly. A proper manufacturing process influences the performance of the door grommet during its lifetime. In this work, uniaxial tensile tests based on molecular dynamics simulations have been performed on an ethylene–propylene–diene monomer (EPDM) material to investigate the effect of the crosslink density and its variation with temperature. The Mooney–Rivlin (MR) model is used to fit the results of molecular dynamics (MD) simulations in this paper and an exponential-type model is proposed to calculate the parameters $C_1\left(T\right)$ and $C_2\left(T\right)$. The experimental results, confirmed by hardness tests of the cured part according to ASTM 1415-88, show that the free volume fraction and the crosslink density have a significant effect on the stiffness of the EPDM material in a deformed state. The results of molecular dynamics superposition on the MR model agree reasonably well with the macroscopically observed mechanical behavior and tensile stress of the EPDM at the molecular level. This work allows the accurate characterization of the stress–strain behavior of rubber-like materials subjected to deformation and can provide valuable information for their widespread application in the injection molding industry.

This paper analyzes Convolutional Neural Network (CNN) models for classifying dried chili pepper quality. The models categorize images into five categories: “Extra”, “First Class”, “Second Class”, “Trash”, and “Empty”, each representing different qualities and scenarios in a sorting machine. We compared architectures from the Torchvision library, including ResNet, ResNeXt, Wide_ResNet, and RegNet using Transfer Learning (TL) in a feature extraction approach. All models employ residual blocks, an innovative technique enhancing deep learning performance. The models were evaluated using crossvalidation and metrics such as Precision, Recall, Specificity, F1-score, Geometric_mean, Index of Balanced Accuracy, and the Matthews Correlation Coefficient. They were trained using SGD, Adagrad, and Adam optimizers. Our findings suggest that ResNet-152, trained with the Adagrad optimizer, achieved the highest mean validation accuracy of 96.62%. The selected model can assist agricultural producers in classifying their products according to international standards.

Convolutional neural networks (CNNs) and digital holographic interferometry (DHI) can be combined to improve the calculation efficiency and to simplify the procedures of many DHI applications. In DHI, for the measurements of concentration differences between liquid samples, two or more holograms are compared to find the difference phases among them, and then to estimate the concentration values. However, liquid samples with high concentration difference values are difficult to calculate using common phase unwrapping methods as they have high spatial frequencies. In this research, a new method to skip the phase unwrapping process in DHI, based on CNNs, is proposed. For this, images acquired by Guerrero-Mendez et al. (Metrology and Measurement Systems 24, 19–26, 2017) were used to train the CNN, and a multiple linear regression algorithm was fitted to estimate the concentration values for liquid samples. In addition, new images were recorded to evaluate the performance of the proposed method. The proposed method reached an accuracy of 0.0731%, and a precision of ±0.0645. The data demonstrated a high repeatability of 0.9986, with an operational range from 0.25 gL⁻¹ to 1.5 gL⁻¹. The proposed method was performed with liquid samples in a cylindrical glass.