Abstract

Abstract
Hexapod robots, with their six-legged design, excel in stability and adaptability on challenging terrain but pose significant control challenges due to their high degrees of freedom. While reinforcement learning (RL) has been explored for robot navigation, few studies have systematically compared on-policy and off-policy methods for multi-legged locomotion. This work presents a comparative study of SARSA and Q-Learning for trajectory control of a simulated hexapod robot, focusing on the influence of learning rate (alpha), discount factor (gamma), and eligibility trace (lambda). The evaluation spans eight initial poses, with performance measured through lateral deviation (Ey), orientation error (E theta), and iteration count. Results show that Q-Learning generally achieves faster convergence and greater stability, particularly with higher gamma and lambda values, while SARSA can achieve competitive accuracy with careful parameter tuning. The findings demonstrate that eligibility traces substantially improve learning precision and provide practical guidelines for robust RL-based control in multi-legged robotic systems.

Abstract
In a high-risk sector, such as power system, transparency and interpretability are key principles for effectively deploying artificial intelligence (AI) in control rooms. Therefore, this paper proposes a novel methodology, the evolving symbolic model (ESM), which is dedicated to generating highly interpretable data-driven models for dynamic security assessment (DSA), namely in system security classification (SC) and the definition of preventive control actions. The ESM uses simulated annealing for a data-driven evolution of a symbolic model template, enabling different cooperative learning schemes between humans and AI. The Madeira Island power system is used to validate the application of the ESM for DSA. The results show that the ESM has a classification accuracy comparable to pruned decision trees (DTs) while boasting higher global inter-pretability. Moreover, the ESM outperforms an operator-defined expert system and an artificial neural network in defining preventive control actions.

Abstract
In recent advancements within sensing technology, driven by the Internet of Things (IoT), significant impacts are observed on health sector applications, notably through wearable electronics like electronic tattoos (e-tattoos). These e-tattoos, designed for direct contact with the skin, facilitate precise monitoring of vital physiological parameters, including body heat, a critical indicator for conditions such as inflammation and infection. Monitoring these indicators can be crucial for early detection of chronic conditions, steering toward proactive healthcare management. This study delves into a thermoelectric sensor e-tattoo designed for detailed skin temperature mapping. Utilizing a novel design, this sensor detects temperature variations across thermoelectric stripes, leveraging screen-printed films of p-type Bi0.35Sb1.65Te3, n-type Bi2Te2.8Se0.2, and poly(vinyl alcohol) (PVA) for enhanced thermoelectric and flexible properties. The application of a prototype printed thermoelectric device on temporary tattoo paper, a pioneering development in wearable health technology is demonstrated. This device, validated through numerical simulations, exhibits significant potential as a non-invasive tool for temperature monitoring, highlighting its value in health diagnostics and management.

Abstract
A notable shift from an internal combustion engine vehicles (ICEVs) fleet to an electric vehicles (EVs) fleet is expected in the medium term due to increasing environmental concerns and technological breakthroughs. In this context, this paper conducts a systematic literature review on life cycle assessment (LCA) research of EVs compared to ICEVs based on highly impactful articles. Several essential aspects and characteristics were identified and discussed, such as the assumed EV types, scales, models, storage technologies, boundaries, lifetime, electricity consumption, driving cycles, combustion fuels, locations, impact assessment methods, and functional units. Furthermore, LCA results in seven environmental impact categories were gathered and evaluated in detail. The research indicates that, on average, battery electric vehicles are superior to ICEVs in terms of greenhouse gas (GHG) emissions (182.9 g CO2-eq/km versus 258.5 g CO2-eq/km), cumulative energy demand (3.2 MJ/km versus 4.1 MJ/km), fossil depletion (49.7 g oil-eq/km versus 84.4 g oil-eq/km), and photochemical oxidant formation (0.47 g NMVOC-eq/km versus 0.61 g NMVOC-eq/km) but are worse than ICEVs in terms of human toxicity (198.1 g 1,4-DCB-eq/km versus 64.8 g 1,4-DCB-eq/km), particulate matter formation (0.32 g PM10-eq/km versus 0.26 g PM10-eq/km), and metal depletion (69.3 g Fe-eq/km versus 19.0 g Fe-eq/km). Emerging technological developments are expected to tip the balance in favor of EVs further. Based on the conducted research, we propose to organize the factors that influence the vehicle life cycle into four groups: user specifications, vehicle specifications, local specifications, and multigroup specifications. Then, a set of improvement opportunities is provided for each of these groups. Therefore, the present paper can contribute to future research and be valuable for decision-makers, such as policymakers.

Abstract
The increasing complexity of wireless environments, driven by user mobility and dynamic obstructions, poses significant challenges to maintaining Line-of-Sight (LoS) connectivity. Mobile base stations (gNBs) offer a promising solution by physically relocating to restore or sustain LoS. This paper explores how reinforcement learning (RL) can be applied to gNB mobility control within vision-aided network systems. As part of the CONVERGE project, we present the CONVERGE Chamber Simulator (CC-SIM), a 3D environment for developing, training, and testing gNB mobility control algorithms. CC-SIM models user and obstacle mobility, visual occlusion, and Radio Frequency (RF) propagation while supporting both offline reinforcement learning and real-time validation through integration with OpenAirInterface (OAI). Leveraging CC-SIM, we trained a Deep Q-Network (DQN) agent that proactively repositions gNBs under dynamic conditions. Across three representative use cases, the agent reduced LoS blockage by up to 42% compared to static deployments, highlighting the potential of RL-driven mobility control and positioning CC-SIM as a practical platform for advancing adaptive, next-generation wireless networks.