Abstract
This work introduces Gen-JEMA, a generative approach based on joint embedding with multimodal alignment (JEMA), to enhance feature extraction in the embedding space and improve the explainability of its predictions. Gen-JEMA addresses these challenges by leveraging multimodal data, including multi-view images and metadata such as process parameters, to learn transferable semantic representations. Gen-JEMA enables more explainable and enriched predictions by learning a decoder from the embedding. This novel co-learning framework, tailored for directed energy deposition (DED), integrates multiple data sources to learn a unified data representation and predict melt pool images from the primary sensor. The proposed approach enables real-time process monitoring using only the primary modality, simplifying hardware requirements and reducing computational overhead. The effectiveness of Gen-JEMA for DED process monitoring was evaluated, focusing on its generalization to downstream tasks such as melt pool geometry prediction and the generation of external melt pool representations using off-axis sensor data. To generate these external representations, autoencoder (AE) and variational autoencoder (VAE) architectures were optimized using Bayesian optimization. The AE outperformed other approaches achieving a 38% improvement in melt pool geometry prediction compared to the baseline and 88% in data generation compared with the VAE. The proposed framework establishes the foundation for integrating multisensor data with metadata through a generative approach, enabling various downstream tasks within the DED domain and achieving a small embedding, allowing efficient process control based on model predictions and embeddings.

Abstract
Wheeled mobile robots are increasingly deployed in harsh environments where dense obstacles, traps, variable terrain, soil effects, tight energy budgets, and sensor noise often deem classical navigation stacks insufficient. This paper presents a PRISMA-guided systematic review of recent work on Deep Reinforcement Learning (DRL) for wheeled ground-robot navigation in harsh environments and organizes the field via a practical six-dimensional taxonomy: environmental challenges, navigation architecture, observation modality, action strategy, action space, and learning algorithm. The taxonomy is refined through an iterative, evidence-grounded coding process on the included studies, and applied under a transparent coding protocol to support reproducible categorization. Across the literature, DRL appears both as a planner module as well as end-to-end policy (behavior) implementer tool. Regarding observation, mapless navigation based on LiDAR or cameras are prevalent. Actions are predicted mostly one time step ahead and are continuous. Actor-critic methods are prevalent, notably PPO and SAC are the common DRL methods used. As for the evaluation methodology, it remains largely simulation-based, with only limited sim-to-real protocols. Building on these findings, we use the previously mentioned taxonomy to identify common design choices for navigation in harsh terrains, propose minimum reporting practices to enable reproducible comparison, and propose research directions including energy-aware learning, improved robustness to sensor degradation, all weather soil-vehicle interaction modeling, short-horizon look-ahead for stability and smoothness, standardized tasks and metrics. The proposed taxonomy and guidelines, as well as identified trends, intend to help researchers and practitioners select methods that best suits their own objectives and constraints, thus hopefully accelerating progress from promising simulation results to dependable, field-ready autonomy.

Abstract
This study examines whether Parameter-Efficient Fine-Tuning (PEFT) of lightweight, free, and open-licensed Large Language Models (LLMs) can yield tutoring assistants for introductory circuit analysis methods, while fitting the students' needs. We analyzed 260 Electrical and Computer Engineering (ECE) exam responses to classify and quantify frequent students' mistakes when applying the Loop Current Method (LCM). Only 28.5% solved the target problem without error, and most difficulties were conceptual (e.g., miscounting the number of independent Kirchhoff's Voltage Law (KVL) equations). Driven by this taxonomy, we assembled official course materials and curated a bilingual (Portuguese-English) pedagogical dataset. Using GTP-4o for distillation, we generated question-answer (QA) pairs for fine-tuning smaller models (Meta Llama 3.2 1B and 3.1 8B), via Quantized Low-Rank Adaptation (QLoRA) on a single commodity GPU, with an end-to-end pipeline completing in under 7 min. A blind study involving 77 first-year ECE students evaluated responses to (never seen) questions from both our tuned models and GPT-4.5, rating correctness, clarity, educational value, task coverage, and style. The 8B model scored within one point (5-point Likert) of GPT-4.5 model and both 1B and 8B were consistently preferred over untuned baseline versions for clarity and task coverage. As a complementary cross-check, 12 higher education senior professors independently evaluated model responses, largely corroborating the student-based rankings. These results provide evidence that carefully curated documents introducing electrical circuit theory, combined with smaller models optimized with PEFT, namely QLoRA, can be used in the construction of a always-available tutoring application. The proposed system features modest cost, runs on consumer-grade hardware, and paves the way for deployable front-end applications that do not involve possibly expensive, resource-hungry, remote machines.

Abstract
Macroeconomic forecasting is a fundamental domain for policy decisions, directly impacting the whole population of a country. The use of machine learning (ML) approaches in economics forecasting has been studied in several types of research in the academic field, aiming to improve or even replace traditional econometric approaches. However, the use of ML in forecasting is now getting closer to policy markers, which are the institutions that make policy decisions. Three relevant studies are presented and analyzed in this work; all focused on forecasting using ML of different macroeconomic variables in several economies. The studies were compared, including aspects of methodologies and results, as well as similarities and differences. In addition, several technical, legal, and philosophical questions were raised regarding the effective use of data from ML forecasting in public policies, including topics related to the standardization of the research on this topic, the explanation of the model's output, protection of trust, and ethics issues.

Abstract
In the context of mobile robotics education, realistic and accessible datasets are fundamental for supporting the development and testing of algorithms. However, collecting real-world data is a limited and challenging task because it is time-consuming and error-prone. Therefore, this paper presents the generation of a synthetic dataset through realistic simulation using the SimTwo environment-a physics-based simulator, and modeling techniques of sensors and actuators. The physical and simulated mobile robot was developed to perform tasks such as following a line, following a wall, and avoiding obstacles. The proposed approach facilitates the creation of customized datasets for training and evaluation algorithms while supporting remote and inclusive learning. Results show that a simulated dataset can effectively replicate real-world behaviors, making them a valuable resource for educational contexts, research, and development. Some emergent machine learning algorithms can be applied to this dataset, being this approach increasingly used to enhance robot localization, by leveraging ML, robots can improve the accuracy, robustness, and adaptability of their localization systems, especially in complex and dynamic environments.

Abstract
The integration of students with intellectual and developmental disabilities into STEAM education presents ongoing challenges, particularly in engineering disciplines where both technical and social competencies are essential. Robotics and active learning methodologies have emerged as promising solutions to address these challenges by offering adaptive, interactive, and student-centered learning environments. This study conducts a systematic literature review to examine how these technologies and methodologies are applied to support students with Intellectual and Developmental Disabilities. A total of 34 high-quality studies published over the past ten years were selected through a rigorous process of database searching, inclusion/exclusion filtering, and quality assessment. The analysis reveals that robotics is particularly effective in fostering academic development, cognitive skills, social-behavioral interaction, and emotional regulation, while active learning promotes social responding, role understanding, and collaborative skills. Together, these approaches not only enhance individual learning outcomes but also facilitate the broader inclusion of students with disabilities within engineering education.