Abstract
Ethanol plays a crucial role in modern industrial processes and consumer products. Despite its presence in human activity, short and long-term exposure to gaseous ethanol poses risks to health conditions and material damage, making the control of its concentration in the atmosphere of high importance. Ethanol optical sensors based on electromagnetic surface waves (ESWs) are presented, with sensitivity to ethanol vapours being achieved by the inclusion of ethanol-adsorptive zinc oxide (ZnO) layers. The changes in optical properties modulate the resonant conditions of ESWs, enabling the tracking of ethanol concentration in the atmosphere. A comprehensive comparative study of sensor performance is carried out between surface plasmon resonance (SPR) and Bloch surface wave (BSW) based sensors. Sensor efficiency is simulated by transfer matrix method towards optimized figures of merit (FoM). Preliminary results validate ethanol sensitivity of BSW based sensor, showcasing a possible alternative to electromagnetic and plasmonic sensors. © The Authors.

Abstract
The potential of digital games, when transformed into Serious Games (SGs), Games for Learning (GLs), or game-based learning (GBL), is truly inspiring. These forms of games hold immense potential as effective learning tools as they have a unique ability to provide challenges that align with learning objectives and adapt to the learner's level. This adaptability empowers educators to create a flexible and customizable learning experience, crucial in acquiring knowledge, experience, and professional skills. However, the lack of a standardised design methodology for challenges that promote skill acquisition often hampers the effectiveness of games-based training. The four-step Triadic Certification Method directly responds to this challenge, although implementing it may require significant resources and expertise and adapting it to different training contexts may be challenging. This method, built on a triadic of components: competencies, mechanics, and training levels, offers a new approach for game designers to create games with embedded in-game assessment towards the certification of competencies. The model combines the competencies defined for each training plan with the challenges designed for the game on a matrix that aligns needs and levels, ensuring a comprehensive and practical learning experience. The practicality of the model is evident in its ability to balance the various components of a certification process. To validate this method, a case study was developed in the context of learning how to drive, supported by a game coupled with a realistic driving simulator. The real time collection of game and training data and its processing, based on predefined settings, learning metrics (performance) and game elements (mechanics and parameterisations), defined by both experts and game designers, makes it possible to visualise the progression of learning and to give visual and auditory feedback to the student on their behaviour. The results demonstrate that it is possible use the data generated by the player and his/her interaction with the game to certify the competencies acquired.

Abstract
The concept of Industry 4.0 and the introduction of the Internet of Things (IoT) on industrial applications, known as Industrial Internet of Things (IIoT), have been changing the scenario of industrial automation. This new paradigm is expected to optimize industrial processes, increase productivity, lower costs and improve operations integration. For that, structured Machine-to-Machine (M2M) communication is key to ensure agility, interoperability and reliability, with several solutions currently available in the literature and in industry. This paper reviews the state of the art on industrial communication protocols and architectures, providing a classification and comparison of these different solutions based on their most relevant features in the context of Industry 4.0. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023.

Abstract
Chest X-ray scans are frequently requested to detect the presence of abnormalities, due to their low-cost and non-invasive nature. The interpretation of these images can be automated to prioritize more urgent exams through deep learning models, but the presence of image artifacts, e.g. lettering, often generates a harmful bias in the classifiers and an increase of false positive results. Consequently, healthcare would benefit from a system that selects the thoracic region of interest prior to deciding whether an image is possibly pathologic. The current work tackles this binary classification exercise, in which an image is either normal or abnormal, using an attention-driven and spatially unsupervised Spatial Transformer Network (STERN), that takes advantage of a novel domain-specific loss to better frame the region of interest. Unlike the state of the art, in which this type of networks is usually employed for image alignment, this work proposes a spatial transformer module that is used specifically for attention, as an alternative to the standard object detection models that typically precede the classifier to crop out the region of interest. In sum, the proposed end-to-end architecture dynamically scales and aligns the input images to maximize the classifier's performance, by selecting the thorax with translation and non-isotropic scaling transformations, and thus eliminating artifacts. Additionally, this paper provides an extensive and objective analysis of the selected regions of interest, by proposing a set of mathematical evaluation metrics. The results indicate that the STERN achieves similar results to using YOLO-cropped images, with reduced computational cost and without the need for localization labels. More specifically, the system is able to distinguish abnormal frontal images from the CheXpert dataset, with a mean AUC of 85.67% -a 2.55% improvement vs. the 0.98% improvement achieved by the YOLO-based counterpart in comparison to a standard baseline classifier. At the same time, the STERN approach requires less than 2/3 of the training parameters, while increasing the inference time per batch in less than 2 ms. Code available via GitHub.

Abstract
The manufacturing domain faces a challenge in making timely decisions due to the large amounts of data generated by digital technologies such as Internet-of-Things, Artificial Intelligence (AI), Digital Twin, and Big Data. By integrating recommendation systems is possible to support the decision-makers in handling large amounts of data by delivering personalised, accurate, and quality recommendations. One example is the SimQL recommendation model that incorporates AI algorithms with trust and similarity measures to enhance recommendation quality. This paper aims to analyse the sensitivity of the SimQL model's parameters, such as dataset conditions, trust and learning factors, and their impact on the final recommendation quality. A fuzzy logic approach is employed to evaluate the model and identify optimal operating conditions for the recommendation system. By implementing the findings of this study, manufacturers can improve the acceptance and adoption of the SimQL trustworthy recommendation system in this field.

Abstract
This work proposes the use of a deep learning-based adversarial diffusion model to address the translation of contrast-enhanced from non-contrast-enhanced computed tomography (CT) images of the heart. The study overcomes challenges in medical image translation by combining concepts from generative adversarial networks (GANs) and diffusion models. Results were evaluated using the Peak signal to noise ratio (PSNR) and structural index similarity (SSIM) to demonstrate the model's effectiveness in generating contrast images while preserving quality and visual similarity. Despite successes, Root Mean Square Error (RMSE) analysis indicates persistent challenges, highlighting the need for continuous improvements. The intersection of GANs and diffusion models promises future advancements, significantly contributing to clinical practice. The table compares CyTran, CycleGAN, and Pix2Pix networks with the proposed model, indicating directions for improvement.