This study focuses on improving the safety of Vulnerable Road Users (VRUs) in traffic by leveraging multispectral imaging and deep learning. We introduce ScaleFuse, a novel RGB-thermal fusion architecture specifically designed to address the limitations of existing multispectral detection methods. Unlike conventional approaches, ScaleFuse performs multiscale feature fusion at intermediate layers of the backbone, adaptively learning spatial and channel-wise importance from both modalities. To ensure robust fusion, we employ the SuperGlue network for precise image alignment, mitigating the common issue of misregistration between RGB and thermal inputs. ScaleFuse is implemented within the YOLOv10 framework, enabling efficient and accurate detection in challenging conditions such as low-light environments, adverse weather, and occlusions. Experimental evaluations on LLVIP, FLIR, and our newly introduced VeNIT-MSD dataset demonstrate that ScaleFuse consistently outperforms single-modality baselines and previous fusion strategies in terms of accuracy, precision, and recall. The proposed system achieves real-time inference while remaining fully compatible with deployment frameworks such as TensorRT, making it suitable for intelligent transportation applications.

Industrial three-phase induction motors are critical to manufacturing operations, yet their failure can cause costly downtime and maintenance overhead. This work introduces a machine learning framework tailored for fault diagnosis and severity assessment in such motors, using synthetic data that replicates six realistic electromagnetic and mechanical fault conditions. The system employs synchronized vibration and current signals, and evaluates several classifiers—including LSTM, GRU, TCN, RNN, and Random Forest—under varying industrial noise profiles such as phase jitter, frequency drift, and transient spikes. Experimental results demonstrate that time-series models consistently outperform classical approaches in noisy environments. For deployment on resource-constrained edge devices, structured pruning and quantization are applied to reduce model size and latency. Beyond motor fault classification, the framework also detects and visualizes approximately spherical clusters to comprehend fault severity levels. The proposed system enables scalable fault monitoring and low-latency predictive maintenance optimized for embedded industrial environments.

DECICE is a Horizon Europe project that is developing an AI-enabled open and portable management framework for automatic and adaptive optimization and deployment of applications in computing continuum encompassing from IoT sensors on the Edge to large-scale Cloud / HPC computing infrastructures. In this paper, we describe the DECICE framework and architecture. Furthermore, we highlight use-cases for framework evaluation: intelligent traffic intersection, magnetic resonance imaging, and emergency response.

Jamming attacks presents a significant challenge to the security and reliability of wireless communication networks, especially in the context of IoT applications. This study introduces a novel approach to detecting jamming attacks by utilizing the upper-layer network parameters from the application and transport layers. An experimental testbed is developed consisting of a Wi-Fi-based IoT server-client application to collect data. The network parameters are gathered from both noiseless and noisy environment conditions to examine the performance variations of different deep-learning models in diverse environments. The performance of various deep learning models is systematically compared, employing evaluation metrics such as accuracy, F1 score, precision, recall, model complexity, and training time. The findings of this research contribute to the development of effective techniques for jamming detection. Moreover, this study provides valuable insights into the selection and adaptation of appropriate models based on system requirements and specifications, enabling efficient detection and mitigation of jamming attacks in wireless communication systems.

The future of goods transportation will rely on increased efficiency, lower risks, and diminished delays through the use of vehicle platoons that benefit from vehicular connectivity using V2X (Vehicle to Everything) applications. This article describes a system that offers the aforementioned vehicular connectivity to platoons, based on AI-enhanced 5G for resource allocation in wireless platoon intra-communications under three scenarios (latency emergency braking, platoon wireless resource management in tunnels, V2X communications interference in a traffic congestion). Demos are described for each of the scenarios, targeting different layers, starting by the PHY (physical) layer where propagation models are implemented, then a simulation-based MAC (medium access control) layer that allows the allocation of resources to the connected User Equipments (UE) and finally a management and orchestration layer capable of monitoring and managing the radio network, offering features such as network slicing management using O-RAN (Open Radio Access Network) standards.

Vehicle platooning is one of the important V2X applications for automated and connected cars. It provides various benefits including low CO2emissions and low fuel consumption in addition to traffic efficiency and safety. However, due the short distances between the vehicles and the obstacles in the environment, the communication reliability becomes of utmost importance for these vehicles for their safe operation. It is essential to analyse platooning system in various conditions. In this work, we present factors and their effects on the communication quality and their effects on the platooning applications. The results obtained from these studies can be used to improve the communication methods during a real-life platooning application for enhanced safety and reliability.

End-of-line (EOL) tests are an important step to detect and respond to reliability issues that electric motors face. In addition to conventional signal processing methods to establish automated test systems, Artificial Intelligence (AI) and Machine Learning (ML) based methods in recent years, managed to become a major enabler for smart manufacturing thanks to advancements in hardware and software components. Inevitably, the importance of quality data made its way into considerations and requirements of automated fault detection and condition monitoring systems. In this regard, this study proposes an AI-based testing framework for electric motors. We provide information on the reasons of faults observed and a test procedure to detect them. We also give detailed specifications on hardware (sensors and data collection equipment), and provide a data architecture and analysis on properties of ML models that make sense to be used in such scenarios.

Building on many discoveries and inventions, electronics started affecting people’s everyday lives in a significant fashion following the invention of the first solid state transistor in late 1940s. The miniaturization paved way for the mass electronics production and later the digital revolution, the outcomes of which are visible to all members of the public today. After about a two-decade-long swing around 2000s from hardware towards software regarding what affects lives more, a point has now been reached where electronics is more important to all and its use is more ubiquitous and crucial than ever before. In most if not all of end user or industrial applications, the capability and quality of electronics hardware are the key determining factors.The European electronics components and systems (ECS) industry has traditionally had a high base line for electronics innovation. However, the industry is now compelled, partly due to competition and partly due customer demand, to manufacture even more reliable electronics products than before. Guaranteeing the reliability of electronics hardware entails the entire ECS value chain to undergo a paradigm shift to holistically address reliability as a key issue. The European ECS industry previously adopted overseas outsourcing considerably, however it is now taking steps to reshape itself into a more coherent value chain with the aim of having not only the electronics designs but also the electronics manufacturing made in Europe.H2020-ECSEL programme successfully funds highly competitive projects in the area of electronics components and systems. We present here a prologue to a similarly funded project entitled Intelligent Reliability 4.0 ("iRel40"), by providing a background to the topic of ECS, project objectives, and the methodologies and implementations we plan to undertake during the 36-month period of this ongoing project.

Understanding vehicle mobility is essential for devising successful protocols for vehicular communications. Vehicles move at varying speeds on roads whose complexities range from plain highway lanes to urban boulevards with intersections/circles, traffic lights and various points of interest on them. This mobility pattern combined with vehicle density fluctuations depending on location and time makes understanding the vehicular mobility a challenging task and leads to vehicular communication solutions which are incomplete or low-performance. In this paper, we propose an alternative method for vehicular mobility analysis. We attempt to extract the spatio-temporal vehicular mobility characteristics of large, complex geographical areas by dividing them into cells and analyzing on these cells certain metrics that are of interest for vehicular communication. We first discuss the logic behind our approach. We then present definitions of the metrics used by the proposed analysis method. We conclude with a demonstration of the results achieved when the proposed method is applied in the analysis of a large scale mobility trace.

Clustering algorithms improve network manageability through several topology partitioning techniques. In some particular cases, such as vehicular ad hoc network (VANETs) communications, significant performance improvements can be introduced via clustered networking solutions whereas merging clusters for the sake of scalability may lead to degraded network stability. In this paper, we explore the impact of merging clusters, and furthermore based on these results, we propose a new clustering technique, namely Relatively Stable Clustering for Unbiased Environments (ReSCUE). The objective of ReSCUE is primarily guaranteeing cluster stability in an unbiased manner. ReSCUE keeps track of the spatio-temporal changes in VANET node characteristics, and uses these characteristics along with local information to prevent biased clustering which is based on common and general node characteristics. We evaluate the performance of ReSCUE through simulations and show that ReSCUE can form relatively more stable clusters while reducing the frequency of cluster merges, as well as that of the node status changes.