Conference Agenda

Digital technologies have begun to represent an important option for improving mental health monitoring, which has become a fundamental challenge worldwide. This study presents a systematic review of the use of wearable sensors, the Internet of Things (IoT), and artificial intelligence (AI) for monitoring and early detection of mental health problems. To this end, 30 studies were analyzed using the PRISMA method and filtered using PICOC criteria. The results show that the most studied disorders are anxiety, stress, and depression. In addition, the most used sensors are those for heart rate, sound, physical activity, and EEG, as they allow for continuous and non-invasive data collection. In terms of AI techniques, most studies use conventional techniques such as SVM, regression, and XGBoost; however, other studies already incorporate more advanced models that combine neural networks and attention mechanisms. These technologies often offer advantages such as early detection, continuous monitoring, and personalized follow-up. However, there are other limitations related to privacy, data variability, and lack of clinical validation. In short, it is concluded that the integration of sensors, AI, and IoT has great potential, but greater standardization and research in diverse populations is required for its real-world application in mental health.

This systematic review analyses the impact of artificial intelligence on the design and improvement of interactive experiences in video games, identifying applied techniques, benefits achieved, and persistent limitations. The study was structured using the PICOC approach, formulating five research questions related to current constraints, AI methods employed, comparison with traditional approaches, specific advantages, and application scenarios. The literature search was conducted in Scopus and external sources using Boolean combinations of terms associated with artificial intelligence and video games. After applying the PRISMA protocol and rigorous inclusion and exclusion criteria, forty studies published between 2021 and 2025 were selected. The results show that AI has a significant impact in five main areas: procedural content generation, reinforcement learning, generative AI, affective computing, and hybrid approaches. These techniques show substantial increases in productivity, player engagement, NPC behavioural diversity, reduced development times, expanded conversational options, and user retention. However, significant challenges remain, particularly related to narrative consistency, content validation, and the partial reduction of designers' creative control. Overall, the findings confirm that the strategic incorporation of AI constitutes a paradigm shift in video game development, rather than an incremental improvement. The future of the sector will depend on hybrid approaches that balance automation and human creativity, as well as longitudinal studies, robust ethical frameworks, and the exploration of applications in emerging environments such as metaverses.

Capturing digital images is currently very simple due to the ease with which one can access a device with a camera, such as digital cameras, tablets, computers, laptops, webcams, and smart devices. Likewise, distributing captured images with or without the owner's authorization is straightforward. In the case of image distribution without the owner's authorization, or, failing that, taking photographs of individuals without their permission, it becomes vital to identify the device from which the images were captured, as this allows the owner of that device to be identified. This work proposes a method for identifying the source of single-tone image capture (images with flat colors and textures). The process consists of a single-tone image (disputed image) whose capture source is to be determined, a set of average images (fingerprint), and a dataset with the fingerprints of the possible devices that could have captured the disputed image. Each fingerprint is composed of a set of images captured by each device, which are used to obtain an average of the statistical distribution of the Photo Response Non-uniformity (PRNU) noise in those images. The comparison between the possible fingerprints and the disputed image is performed by computing the similarity between the two variables using the Mahalanobis distance. The results obtained indicate effective identification of the device that captured the disputed image.

This research aimed to analyze the implementation of the Semantic Web in digital organizations located in Metropolitan Lima, considering organizational factors, implemented technologies, and information technology (IT) strategies associated with this process. A quantitative approach was adopted, using a non-experimental, cross-sectional, descriptive–correlational research design. Data were collected through the analysis of institutional websites and a structured questionnaire administered to IT managers from organizations in the technology, services, and e-commerce sectors. The results showed that the implementation of the Semantic Web was at an early stage, characterized by low levels of adoption and limited use of semantic technologies such as RDF, OWL, and SPARQL. Additionally, factors such as lack of time, budget constraints, and uncertainty regarding return on investment significantly influenced adoption levels. Conversely, organizations with IT strategies aligned with organizational objectives and an innovation-oriented approach exhibited higher levels of implementation. It was concluded that the effective adoption of the Semantic Web in digital organizations in Metropolitan Lima depends not only on technological availability but also on the strengthening of IT strategies and the mitigation of organizational and economic barriers.

Bamboo is a renewable and abundant material, but its geometric variability (diameter, ovality, taper, wall thickness) complicates the development of standardized connections for structural applications. This paper presents a parametric CAD/CAM toolchain, implemented in Rhino/Grasshopper, that generates adaptive steel joints for bamboo culms of varying sizes and connection angles. The system encodes geometric constraints and structural rules, particularly the orientation of steel plates according to section properties, producing fabrication-ready outputs (DXF and assembly guides) suitable for manual or CNC manufacturing. A prototype joint was fabricated and tested in a cantilever bending experiment with a 3-m bamboo member. Results demonstrated a linear load–deflection response up to 39.2 N, with failure at 49 N due to bamboo splitting. The implemented orientation rules improved bending resistance by a factor of ≈50×, confirming the role of simple codified decisions in enhancing performance. The contribution aligns with ICITS themes on software systems (D), systems modeling (C), and decision support (G), illustrating how digital fabrication can translate computational design into practical, low-cost technologies for sustainable rural construction.

There is a need to monitor individuals performing critical tasks who are affected by drowsiness or fatigue using computer vision detection processes on embedded devices, but this presents a challenge due to the use of complex processing hardware. Previous research describes how embedded devices and microcontrollers perform drowsiness detection tasks using algorithms optimized for limited resources. Additionally, facial recognition methods, such as Haar Cascades or convolutional neural networks, are used to identify fatigue based on facial features. This research implements a computer vision-based drowsiness detection system that calculates facial metrics on a Raspberry Pi 4, using MediaPipe and OpenCV tools, integrating an ESP32 module for MQTT alerts, and a remote monitoring platform with Node-RED. The system recorded an average processing time per frame of 129.08 ms, with increases during drowsiness events (eye closure at 270 ms and yawning at 230 ms). Furthermore, the Raspberry Pi analysis showed an average CPU usage of 51%, indicating its stability despite variations in computational load. The results indicate that it is possible to detect drowsiness and yawning states, characterizing their computational performance, although it has limitations due to the camera resolution and processing time.