The healthcare sector is experiencing a profound transformation, fueled by the rapid evolution of sixth-generation (6G) cellular networks and Internet of Things (IoT) technologies. At the heart of this shift lies the concept of medical digital twins (MDTs), which serve as dynamic virtual representations of physical systems or biological processes. MDTs offer a secure environment to simulate and evaluate therapeutic strategies, leading to reduced costs and more informed clinical decision-making. They also enable real-time support and in-depth data analysis, setting new standards for patient care. Nonetheless, realizing the full capabilities of MDTs remains challenging due to the inherent complexity of human life cycles. Crucial aspects include selecting appropriate data sources and defining robust communication protocols between the physical and digital realms. In particular, integrating wearable technologies with edge computing and WiFi-based Channel State Information (CSI) can significantly enhance health monitoring and activity recognition for elderly individuals within indoor settings. The synergy of IoT advancements and 6G networks paves the way for improved data exchange and continuous synchronization between digital and physical counterparts. This paper, part of the HIPPOCRATES project, presents an IoT-driven architecture for MDTs that incorporates wearable sensors and CSI data to strengthen health monitoring and early intervention strategies, with a focus on elderly care.

Climate change mitigation requires the urgent reduction of Greenhouse Gas (GHG) emissions, with the building sector as a significant contributor. This study develops a system to identify appliance profiles from smart meter data, enhancing energy consumption awareness and management. These profiles provide valuable insights into users’ consumption patterns and habits, enabling more accurate load consumption prediction and effective appliance scheduling strategies. The proposed approach employs feature extraction techniques to characterise energy consumption profiles, followed by k-means clustering to identify distinct appliance profiles. Eleven representative features are identified, offering comprehensive insights into occupants’ energy usage habits. The evaluation with real-case data shows accurate consumption cycle approximations for each profile, with errors consistently below 10%. Performance assessment using classification metrics indicates well-characterised and representative profiles, outperforming state-of-the-art methods with average values exceeding 0.88 for all considered metrics. This system helps raise occupants’ awareness of appliance energy usage and facilitates optimised scheduling through an Energy Management System (EMS). By promoting more efficient energy consumption, the proposed approach contributes to overall energy reduction and, consequently, lower GHG emissions in the building sector. • Development of a system to monitor consumption habits for different appliances. • Design a methodology to identify key features for diverse appliance profiles. • Identification of 11 robust features representing consumption profiles effectively. • Integration of k-means clustering to group appliances by consumption and behaviour. • Validation of the proposed system using a real-case dataset.

Human Activity Recognition (HAR) techniques play a key role in identifying and categorizing human activities based on environmental information. More recently, the use of Channel State Information (CSI) has gained momentum because this information can be extracted in a non-intrusive manner. CSI-based algorithms leverage the correlation between CSI dynamics of wireless transmissions and human body movements. However, privacy concerns may arise, as this approach may inadvertently disclose sensitive information about individuals’ movements, habits, and behaviors. In this context, this study investigates the challenge of preserving user privacy in CSI-based HAR for eHealth Ambient Assisted Living (AAL) applications. More specifically, the impact of simple filter-based CSI obfuscation is evaluated on the accuracy performance of a HAR model that makes use of a Long-Short-Term Memory (LSTM) algorithm. Using a publicly available dataset, the accuracy between the original and the obfuscated versions of the dataset generated by simple filtering techniques is compared. The results show significant performance degradation when data is obfuscated, with a HAR accuracy degradation compared to the original results ranging from a minimum of 17.9% to a maximum of nearly 90%. Such results prove that, even with simple obfuscation techniques, the privacy of CSI-enabled HAR-based systems can be sufficiently preserved.

This paper addresses the problem of data privacy in Human Activity Recognition (HAR) applications for eHealth. Cryptography, a proven privacy safeguard on the Internet, remains underutilized in the HAR context, as observed in the existing literature. This study highlights the importance of cryptographic practices by conducting a performance analysis, focusing on the accuracy of HAR with and without data en-cryption. This paper proves that even by introducing a very simple cryptographic mechanism, a potential eavesdropper would experience a reduction of more than 20% of accuracy in the activity recognition task as compared to the case where no encryption is used, at the price of a limited increase in the energy consumption for the involved sensors. Such preliminary results demonstrate the effectiveness of encryption for applications of this type, encouraging further exploration and refinement in this direction.

It has been largely demonstrated how human be-haviour can have a great impact on the quality of life. Some habits, such as those related to sleeping, exercising and working, directly affect people's psycho-physical health, either positively or negatively. The research has been increasingly focusing on better understanding how some behaviours, or changes in someone's usual habits, can be triggers to early recognising and predicting bad health conditions that might even be the warning signal of pathological conditions such as depressive disorders or neurodegenerative diseases. Non-invasive wearable health monitoring systems have the potential of being a key technology to this purpose, because they are easy-to-use and are not perceived as intrusive by users. In this paper, a system that makes use of popular commercial wrist-wearable devices to find the correlation between the monitored users' activities and their stress and well-being conditions, as subjectively self-assessed by them, is proposed. The paper aims to present a methodology to automatically learn which users' activities can be associated with positive and negative health conditions so that they can be later predicted as soon as the first signals are detected by wearable devices. The paper further presents the implementation and preliminary results of a first prototype of the proposed system, which monitors users' sleep and activity and assesses the correlation with stress levels and illness conditions.