This paper examines the recent advances made in the field of Deep Learning (DL) methods for the automated identification of Partial Discharges (PD). PD activity is an indication of the state and operational conditions of electrical equipment systems. There are several techniques for on-line PD measurements, but the typical classification and recognition method is made off-line and involves an expert manually extracting appropriate features from raw data and then using these to diagnose PD type and severity. Many methods have been developed over the years, so that the appropriate features expertly extracted are used as input for Machine Learning (ML) algorithms. More recently, with the developments in computation and data storage, DL methods have been used for automated features extraction and classification. Several contributions have demonstrated that Deep Neural Networks (DNN) have better accuracy than the typical ML methods providing more efficient automated identification techniques. However, improvements could be made regarding the general applicability of the method, the data acquisition, and the optimal DNN structure.

This paper aims to design and implement a system capable of distinguishing between different activities carried out during a tennis match. The goal is to achieve the correct classification of a set of tennis strokes. The system must exhibit robustness to the variability of the height, age or sex of any subject that performs the actions. A new database is developed to meet this objective. The system is based on two sensor nodes using Bluetooth Low Energy (BLE) wireless technology to communicate with a PC that acts as a central device to collect the information received by the sensors. The data provided by these sensors are processed to calculate their spectrograms. Through the application of innovative deep learning techniques with semi-supervised training, it is possible to carry out the extraction of characteristics and the classification of activities. Preliminary results obtained with a data set of eight players, four women and four men have shown that our approach is able to address the problem of the diversity of human constitutions, weight and sex of different players, providing accuracy greater than 96.5% to recognize the tennis strokes of a new player never seen before by the system.

Counting has become a fundamental task for data processing in areas such as microbiology, medicine, agriculture and astrophysics. The proposed SA-CNN-DC (Scale Adaptive—Convolutional Neural Network—Distance Clustering) methodology in this paper is designed for automated counting of steel bars from images. Its design consists of two Machine Learning techniques: Neural Networks and Clustering. The system has been trained to count round and squared steel bars, obtaining an average detection accuracy of 98.81% and 98.57%, respectively. In the steel industry, counting steel bars is a time consuming task which highly relies on human labour and is prone to errors. Reduction of counting time and resources, safety and productivity of employees and high confidence of the inventory are some of the advantages of the proposed methodology in a steel warehouse.

Household appliances, climate control machines, vehicles, elevators, cash counting machines, etc., are complex machines with key contributions to the smart city. Those devices have limited memory and processing power, but they are not just actuators; they embed tens of sensors and actuators managed by several microcontrollers and microprocessors communicated by control buses. On the other hand, predictive maintenance and the capability of identifying failures to avoid greater damage of machines is becoming a topic of great relevance in Industry 4.0, and the large amount of data to be processed is a concern. This article proposes a layered methodology to enable complex machines with automatic fault detection or predictive maintenance. It presents a layered structure to perform the collection, filtering and extraction of indicators, along with their processing. The aim is to reduce the amount of data to work with, and to optimize them by generating indicators that concentrate the information provided by data. To test its applicability, a prototype of a cash counting machine has been used. With this prototype, different failure cases have been simulated by introducing defective elements. After the extraction of the indicators, using the Kullback–Liebler divergence, it has been possible to visualize the differences between the data associated with normal and failure operation. Subsequently, using a neural network, good results have been obtained, being able to correctly classify the failure in 90% of the cases. The result of this application demonstrates the proper functioning of the proposed approach in complex machines.

Smart IoT sensors are characterized by their ability to sense and process signals, producing high-level information that is usually sent wirelessly while minimising energy consumption and maximising communication efficiency. Systems are getting smarter, meaning that they are providing ever richer information from the same raw data. This increasing intelligence can occur at various levels, including in the sensor itself, at the edge, and in the cloud. As sending one byte of data is several orders of magnitude more energy-expensive than processing it, data must be handled as near as possible to its generation. Thus, the intelligence should be located in the sensor; nevertheless, it is not always possible to do so because real data is not always available for designing the algorithms or the hardware capacity is limited. Smart devices detecting data coming from inertial sensors are a good example of this. They generate hundreds of bytes per second (100 Hz, 12-bit sampling of a triaxial accelerometer) but useful information comes out in just a few bytes per minute (number of steps, type of activity, and so forth). We propose a lossy compression method to reduce the dimensionality of raw data from accelerometers, gyroscopes, and magnetometers, while maintaining a high quality of information in the reconstructed signal coming from an embedded device. The implemented method uses an adaptive vector-quantisation algorithm that represents the input data with a limited set of codewords. The adaptive process generates a codebook that evolves to become highly specific for the input data, while providing high compression rates. The codebook’s reconstruction quality is measured with a peak signal-to-noise ratio (PSNR) above 40 dB for a 12-bit representation.

This study presents a novel system for diagnosing and evaluating soccer performance using wearable inertial sensors integrated into players’ insoles. Designed to meet the needs of professional podiatrists and sports practitioners, the system focuses on three key soccer-related movements: passing, shooting, and changes of direction (CoDs). The system leverages low-power IMU sensors, Bluetooth Low Energy (BLE) communication, and a cloud-based architecture to enable real-time data analysis and performance feedback. Data were collected from nine professional players from the SD Huesca women’s team during controlled tests, and bespoke algorithms were developed to process kinematic data for precise event detection. Results indicate high accuracy rates for detecting ball-striking events and CoDs, with improvements in algorithm performance achieved through adaptive thresholds and ensemble neural network models. Compared to existing systems, this approach significantly reduces costs and enhances practicality by minimizing the number of sensors required while ensuring real-time evaluation capabilities. However, the study is limited by a small sample size, which restricts generalizability. Future research will aim to expand the dataset, include diverse sports, and integrate additional sensors for broader applications. This system offers a valuable tool for injury prevention, player rehabilitation, and performance optimization in professional soccer, bridging technical advancements with practical applications in sports science.