Abstract
This work introduces JEMA (Joint Embedding with Multimodal and multi-view Alignment), a novel co-learning framework and loss function to combine multiple sensors and process parameters in Directed Energy Deposition (DED), a critical process in metal additive manufacturing. As Industry 5.0 advances in industrial applications, effective process monitoring becomes increasingly essential. However, the limited availability of data and the black-box nature of AI solutions present significant implementation challenges in industrial settings. JEMA addresses these limitations by leveraging multimodal data, including multi-view images and process parameters, to learn transferable semantic representations. By implementing a supervised regression contrastive loss function, JEMA shapes the embedding space to enable interpretable inference. Furthermore, the framework allows for simplified hardware requirements and reduced computational overhead during deployment by utilizing only the primary on-axis sensor. We evaluate the effectiveness of JEMA loss in DED process monitoring, with particular focus on its generalization capabilities for downstream tasks such as melt pool geometry prediction without extensive fine-tuning. Our empirical results demonstrate the effectiveness of JEMA, showing improvements of 29% and 20% in multimodal and unimodal settings, respectively, compared to models without any regularization loss. Additionally, JEMA outperforms supervised contrastive learning methods by 8% and 2% in the same settings. These improvements are also accompanied by a more structured and meaningful representation in the embedding space. Importantly, the learned embedding representation provides direct interpretability of the feature space, which can be utilized by both human operators and automated systems for process optimization, control, and anomaly detection based on defined thresholds. This human-centered approach ensures that operators can actively engage with the system, making informed decisions and enhancing their trust in the process. Our framework establishes a foundation for integrating multisensor data with metadata, enabling diverse downstream applications both within manufacturing processes and beyond, while keeping human expertise central to the loop.

Abstract

Abstract

Abstract
This chapter explores the innovative development and integration of tailored user research instruments to inform digital health solutions for People Living with Amyotrophic Lateral Sclerosis (PALS) exhibiting characteristics of partial Locked-In Syndrome (LIS). Addressing the complex interplay of motor, cognitive, and emotional impairments typical of this population, the study proposes a synergistic framework combining three adapted instruments: the ALS Functional Rating Scale-Revised (ALSFRS-R/EX), the User Experience Questionnaire Plus (UEQ+), and a bespoke Cognitive-Motor-Emotional (CME) Observation Grid. These instruments were tailored to detect subtle variations in user function, affect, and interaction. Results show how embodied and sensory drawing participatory methods and customisation of instruments, along with semi-structured questionnaire and interviews with caregivers, can yield actionable insights for designing a model for solutions in neurodegenerative or communication-limiting contexts beyond Augmentative and Alternative Communication (AAC).

Abstract
This study explores the complex process of entrepreneurial team identity formation and development, addressing a notable gap in the current literature. Focusing on five entrepreneurial teams affiliated with CEU iLab, the study adopts a multiple case study design drawing on semi-structured interviews with program alumni, complemented by secondary data obtained through manual web scraping. Findings reveal that entrepreneurial identity begins forming even before teams enter the incubation program and evolves through a dynamic interplay of factors. High levels of social interaction and networking, team stability, intra-team trust, effective feedback mechanisms, and perceived legitimacy all contribute to shaping this identity. The incubation setting acts as a catalyst, reinforcing these mechanisms and accelerating identity development. This research offers theoretical contributions by proposing a model of entrepreneurial team identity formation and highlighting how relational and contextual factors influence this ongoing process.

Abstract
Brain connectivity analysis based on stereo-electroencephalography (SEEG) has been increasingly explored as a tool for studying Epileptogenic Networks and supporting therapies for Refractory Epilepsy. In this Review, we summarize recent methodologies and findings on SEEG-based Functional and Effective Connectivity, which aim to define the dependence and influence between spatially distributed neurophysiological events. After briefly outlining the network analysis measurements used to interpret connectivity, we detail the mathematical principles underlying the 11 reviewed connectivity methods, along with the main findings characterizing the epileptogenic network for each method. Based on their principles, these methods can be organized into four classes: Signal Interaction, Signal Synchronization, Causality, and Other Nonlinear methods. Across these methods, consistent network signatures emerge, including functional isolation of the Epileptogenic Zone during interictal periods and complex hypersynchronization dynamics during seizures. As there is no single best method, we present a methodological framework that summarizes the relationship assumptions, computational costs, signal-processing requirements, and clinical use cases for each method. Finally, we discuss current challenges and future trends in this field.