Currently used recycling technologies have limitations on the composition of recyclable waste, which makes them specialized. Thus, the preliminary sorting of municipal solid waste is a necessary step, increasing the efficiency of using municipal solid waste as a resource. To sort municipal solid waste we developed a method for detecting and classifying waste on a conveyor line using neural network image processing. Images from a camera are fed to a neural network input, which determines the position and type of detected objects. To train the neural network a database of more than 13,000 municipal solid waste images was created. Mean-Average Precision for the neural network model was 64%.

Graph neural networks (GNNs) represent a promising instrument for surrogate modeling, capable of handling unstructured computational meshes naturally. We address a typical issue of the accuracy degradation for larger computational domains due to the limited receptive field of GNN models and long-range global interactions between nodes of the mesh. We propose a modification of the GNN architecture allowing to improve the accuracy by a factor of 3 without significant increase in computational costs. The validation tests of the model concentrate on the two-dimensional stationary fluid flow around a bluff body in a channel and corresponding heat transfer. The problem formulation includes bluff bodies of randomly generated shapes and various boundary conditions. The model shows a robust performance for the out-of-domain data, i.e., the flow over an airfoil for different angles of attack.

The application of machine learning to solve engineering problems is in extremely high demand. This article proposes a tool that employs machine learning algorithms for predicting the frequency response of an unsteady vortex phenomenon, the precessing vortex core (PVC), occurring in a conical diffuser behind a radial swirler. The model input parameters are the two components of the time-averaged velocity profile at the cone diffuser inlet. An empirical database was obtained using a fully automated experiment. The database associates multiple inlet velocity profiles with pressure pulsations measured in the cone diffuser, which are caused by the PVC in the swirling flow. In total, over 10³ different flow regimes were measured by varying the swirl number and the cone angle of the diffuser. Pressure pulsations induced by the PVC were detected using two pressure fluctuations sensors residing on opposite sides of the conical diffuser. A classifier was constructed using the Linear Support Vector Classification (Linear SVC) model and the experimental data. The classifier based on the average velocity profiles at the cone diffuser inlet allows one to predict the emergence of the PVC with high accuracy (99%). By training a regression artificial neural network, the frequency response of the flow was predicted with an error of no more than 1.01 and 5.4% for the frequency and power of pressure pulsations, respectively.

Radio detection of air showers produced by ultra-high energy cosmic rays is a cost-effective technique for the next generation of sparse arrays. The performance of this technique strongly depends on the environmental background, which has different constituents, namely anthropogenic radio frequency interferences, synchrotron galactic radiation and others. These components have recognizable features, which can help for background suppression. A powerful method for handling this is the application of convolution neural networks with a specific architecture called autoencoder. By suppressing unwanted signatures, the autoencoder keeps the signal-like ones. We have successfully developed and trained an autoencoder, which is now applied to the data from Tunka-Rex. We show the procedures of the training and optimization of the network including benchmarks of different architectures. Using the autoencoder, we improved the standard analysis of Tunka-Rex in order to lower the threshold of the detection. This enables the reconstructing of sub-threshold events with energies lower than 0.1 EeV with satisfactory angular and energy resolutions.

We apply deep learning algorithms to tackle the bubble recognition task relying on the experimental video recordings of the vapor cavities growing during the water pool boiling due to the heated bottom and an isothermal multiphase flow in a channel. As a basic network architecture we use U-Net with ResNet 34 and ResNet 50 encoders depending on the complexity of the image background. Three classes have been introduced, i.e. the background, bubble and its boundary allowing to post-process some geometric characteristics in a straightforward manner. We demonstrate the capabilities by tracking the growth of an ensemble of vapor bubbles attached to the heater and studying the size distribution of bubbles in a channel.

District heating networks (DHNs) are critical infrastructure prone to physical failures such as leakage-related faults, which cause significant energy and financial losses. Traditional physics-based monitoring methods are computationally expensive and require the continual recalibration of complex mathematical models, while standard data-driven approaches often fail due to the scarcity of real-world sensor data. This study addresses these challenges by proposing a topology-aware graph neural network (GNN) architecture for fault localization. The methodology follows a two-stage process: first, a graph attention-based architecture is designed and optimized using a synthetic dataset to effectively capture multi-step neighborhood dependencies. Second, the model is adapted and evaluated on a physically simulated dataset of a real urban DHN, comprising 187 nodes and 42,570 operational states. The problem is formulated as a multi-class classification task across supply and return subnets. The results demonstrate high predictive performance, achieving an accuracy of 96% on the supply subnet and 91% on the return subnet. Analysis of prediction errors reveals a strong bias towards local topological mistakes, indicating the model’s ability to capture the physical propagation of disturbances. These findings highlight the efficacy of GNNs in handling sparse data and exploiting network topology for robust DHN monitoring.