Abstract

Abstract
Group fairness in machine learning is an important area of research focused on achieving equitable outcomes across different groups defined by sensitive attributes such as race or gender. Federated learning, a decentralized approach to training machine learning models across multiple clients, amplifies the need for fairness methodologies due to its inherent heterogeneous data distributions that can exacerbate biases. The intersection of federated learning and group fairness has attracted significant interest, with 48 research works specifically dedicated to addressing this issue. However, no comprehensive survey has specifically focused on group fairness in Federated Learning. In this work, we analyze the key challenges of this topic, propose practices for its identification and benchmarking, and create a novel taxonomy based on criteria such as data partitioning, location, and strategy. Furthermore, we analyze broader concerns, review how different approaches handle the complexities of various sensitive attributes, examine common datasets and applications, and discuss the ethical, legal, and policy implications of group fairness in FL. We conclude by highlighting key areas for future research, emphasizing the need for more methods to address the complexities of achieving group fairness in federated systems.

Abstract
Two-photon interference in a Hong-Ou-Mandel (HOM) interferometer can be used as a quantum sensing mechanism due to the sensitivity of the interference dip to perturbations of the photon indistinguishability. In particular, recent studies have generalized this concept to microscopy setups, but the sensitivity to optical path differences constrains its application to samples with thickness variation typically below a few micrometers if tracking changes in the coincidences at a fixed delay. Extending the concept to polarization microscopy and circumventing this limitation, this manuscript explores the use of a narrowband photon pair source with coherence length >1 mm to broaden the HOM dip. Thus, realistic sample-thickness variations introduce negligible temporal distinguishability, and changes in coincidence rate at the dip center are then dominated by sample-induced polarization effects. To compute the polarization rotation, we develop a statistical model for the interferometer, derive the Fisher information, and establish a maximum-likelihood estimator for the local fast-axis angle. Recording dip and baseline frames at each sample position via raster scanning, the experimental results validate the framework, agreeing with classical polarized-intensity images while demonstrating operation close to the maximum-precision regime allowed by the statistical model and insensitivity to layer thickness. Overall, this proof-of-concept demonstrates thickness-insensitive quantum mapping of the local fast-axis angle, which may motivate future developments toward more comprehensive birefringence characterization and low-damage imaging of photosensitive samples, while the present proof-of-concept focuses on the half-wave-like regime, where the local fast-axis angle can be estimated with a single-parameter model.

Abstract
The increasing deployment of solar photovoltaic technologies has intensified concerns regarding end-of-life waste management and the recovery of critical raw materials. Given the socio-environmental and economic significance of photovoltaic panels’ waste, their integration into circular economy and industrial symbiosis strategies is becoming imperative. However, current sustainability assessment methods remain fragmented, with few frameworks adequately supporting informed decision-making across sustainability dimensions. This study conducts a literature review of existing decision-support frameworks that integrate Life Cycle Assessment and Multi-criteria Decision Analysis in the context of industrial symbiosis. The results reveal limited applications of this hybrid methodology specifically targeting photovoltaic waste streams. Key challenges, gaps, and trends were identified being shared into particular inputs and holistic outputs. Based on this synthesis, the paper proposes foundational features for a robust framework tailored to the industrial symbiosis of the photovoltaic waste context, emphasizing dynamic modeling and the critical role of digital tools. This work contributes both a conceptual roadmap and a practical foundation for researchers, policymakers, and industry actors seeking to enhance the sustainability and circularity of photovoltaic waste through industrial symbiosis. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2026.

Abstract
Prolonged standing and repetitive lifting are routine occupational stressors that elevate plantar pressures across workers. In those with diabetes, these demands represent additional risk factors for diabetic foot pathology, highlighting the need for ergonomic interventions beyond standard safety footwear. This study evaluated the perceived ergonomic performance of the MoonWalking (R) insole, a novel adaptive pneumatic system designed for real-time pressure stabilization and offloading when integrated into safety footwear. A comparative experimental protocol tested two conditions: safety footwear with the manufacturer's original insole and the same footwear with the MoonWalking prototype. Twenty participants assessed perceived comfort using a VAS and binary ergonomic questionnaires. The results showed statistically significant improvements in perceived cushioning, foot fit, and overall comfort when using the MoonWalking insole. Participants consistently identified pressure-stabilizing and offloading functions across all plantar regions, indicating that adaptive pressure control was clearly perceptible. No pain or movement restrictions were reported. Although perceived fatigue did not reach statistical significance, a decreasing trend was observed. A slight reduction in intention to reuse the footwear occurred with the prototype, possibly due to its increased weight. These findings provide evidence that integrating an adaptive pneumatic insole into safety footwear may improve plantar pressure redistribution and user comfort.

Abstract
PurposeDetecting anomalous access to electronic health records (EHRs) is critical for safeguarding patient privacy and ensuring compliance with healthcare regulations. Traditional anomaly detection methods often struggle in this domain due to extreme class imbalance, limited labelled data and the subtlety of insider threats. This study proposes a lightweight, hybrid anomaly detection framework that integrates unsupervised, supervised and rule-based approaches using a meta-classifier architecture.Design/methodology/approachAn experimental and model-development approach is employed, combining machine learning techniques with domain-inspired rule modelling to construct a hybrid anomaly detection framework for healthcare access logs. Performance of the algorithm is measured using standard classification metrics such as precision, recall, F1-score and accuracy.FindingsEvaluated on a synthetic but realistic dataset of 50.000 normal and 500 labelled anomalous healthcare access events, the proposed framework achieved superior performance compared to standalone models as well as other hybrid models, with an F1-score of 0.8989 and recall of 0.8180. It also maintained low inference latency (0.028 ms) and energy consumption (4.03e-07 kg CO2), making it suitable for deployment in resource-constrained clinical environments.Originality/valueThis study highlights the potential of a hybrid meta-classifier to enhance anomaly detection in healthcare access logs, capturing both subtle and obvious anomalies while outperforming conventional models and remaining efficient, scalable and practical for real-time monitoring.