Technological Institute of the Philippines

Introduction. Burn injuries are a common source of preventable morbidity in low- and middle-income countries (LMICs). In many communities, burn first aid is influenced by generational traditions and locally available topical remedies, which can replace or delay effective cooling with water recommended in evidence-based guidance. Programme Approach. This brief report presents Burn Prevention and Treatment Education (BPTE), a culturally adapted burn first aid education programme implemented through a community clinic in a Southeast Asian LMIC. BPTE was intentionally designed as a low-technology intervention for contexts with limited electricity and variable access to running water, using short teaching scripts, visual aids, and repeated reinforcement strategies. Implementation Observations. The programme core message (“Cool, Call, Cover”) was refined through iterative adaptations to improve feasibility and comprehension. Adaptations included using hand water pump imagery to represent water access, practical water collection guidance to address contamination concerns, and continuous clinic-based reinforcement through large printed tarpaulins displayed at the clinic gate. Conclusions. BPTE illustrates how evidence-based burn first aid education can be translated into feasible, culturally acceptable actions in resource-limited settings. While clinical endpoints were not systematically collected, the report provides implementation-focused insights into barriers, adaptation strategies, and sustainable delivery channels. Future evaluation should integrate structured documentation of first aid actions and escalation patterns to support impact measurement.

The education sector is mandated to contribute to a quality nation capable of transcending the social, political, economic, cultural, and ethical issues that require human development, productivity, and global competitiveness. Globalization brought about a worldwide flow of ideas, resources, people, economy, values, culture, knowledge, goods, services, and technology. In the Philippines, several local government units have taken the initiative to make higher education accessible to their constituents by establishing through ordinances of their respective Sanggunians, local colleges, and universities (LCUs) within their territorial jurisdiction under the Local Government Code of 1991. The rationale for establishing LUCs points towards the democratizing effects of having local higher learning institutions that are accessible to residents and affordable to financially challenged constituents. The LGUs must make sure that the higher education institutions they have opened will be able to deliver quality education or offer courses according to the standards set by the Commission on Higher Education. The LGU must make sure that it has the available resources to fund the efficient operation of the school, i.e., the building and maintenance of the educational facilities, program offerings, human resources development, and accreditation—all these functions need budget allocations. This study employed descriptive and correlational research methods. Descriptive research characterizes the nature and behavior of the interaction, event, and phenomena resulting from the use of a survey through the descriptive data collection technique.

This paper presents an adaptive tuning framework for Unscented Kalman Filtering (UKF) in Inertial Measurement Unit–Global Navigation Satellite System (IMU–GNSS) vehicle localization using Particle Swarm Optimization (PSO) across simulation and on-road validation. In the first stage, the PSO-tuned UKF is evaluated in the CARLA 0.9.14 simulator with a Tesla Model 3 under various maneuvers and environments. PSO jointly optimizes the scaling parameters α,β,κ and noise covariances Q and R. Simulation results show that the PSO-tuned UKF reduces root-mean-square error (RMSE) to 0.50 m, an 82.14% improvement over a manually tuned UKF and a substantial mitigation of IMU-only drift, while maintaining 1.35 ms update time at 100 Hz. In the second stage, an instrumented vehicle with automotive-grade IMU, GPS, and wheel-speed sensing provides logs for offline replay, PSO re-tuning on field data, and embedded deployment, yielding hypothetical on-road RMSE of 0.72 m versus 1.18 m for manual tuning. This two-stage framework bridges CARLA-based optimization and real-time IMU–GPS localization in on-road environments with statistically validated performance.

This study explores the adoption of a hybrid learning model that integrates Augmented Reality (AR) with a Learning Management System (LMS) to enhance STEM (Science, Technology, Engineering, and Mathematics) education in Philippine schools. Aimed at addressing long-standing challenges such as limited access to quality instructional materials, insufficient teacher training, and low student engagement, the paper introduces an immersive AR-enhanced learning kit embedded within an LMS platform. The hybrid approach seeks to deliver interactive, personalized, and scalable e-learning experiences that align with the 21st-century educational needs. Using thematic analysis of qualitative data from 263 STEM educators in the Greater Manila Area, the study identifies strong institutional support for AR-LMS adoption alongside key barriers, including infrastructural limitations and cost concerns. Findings highlight that integrating AR into LMS-supported instruction significantly improves conceptual understanding, learner motivation, and instructional engagement in STEM-related subjects. The research underscores the potential of AR-enabled e-learning tools to transform traditional teaching methodologies, promote active learning, and support digital transformation in education. The study offers a scalable and inclusive framework for enhancing the STEM competencies of learners in the country, and it perfectly aligns with global educational innovation goals.

This qualitative case study investigates the risk management strategies employed by Ramelex Pharma Corporation, a pharmaceutical distribution firm in the Philippines. The study focuses on financial, operational, and reputational risks through thematic analysis of interviews with key decision-makers. Insights reveal that financial risks are directly linked to employee welfare and operational resilience; operational risks are critical to ensuring public health; and reputational risks significantly influence stakeholder trust and business continuity. The findings validate that an integrated, leadership-driven approach grounded in contextual realities is essential for effective risk management in the pharmaceutical sector. Practical strategies and leadership responses highlighted in the study offer guidance for building resilient operations in similar high-stakes industries, particularly in the digital era of Industry 4.0.

Intercropping is important for sustainable agriculture, especially in Sibuyan Island, Philippines, where fruit trees, timber trees, and crops are grown together. Hence, the lack of scalable and accurate plant identification methods hinders production in such systems. This research proposes leaf-based classification of 29 crop and tree species utilizing a unique dataset of 5,627 images from natural intercropping settings and EfficientNet, a state-of-the-art convolutional neural network. Four EfficientNet versions (EfficientNet224- EfficientNet300) were trained and assessed for accuracy, precision, recall, and F1-score. EfficientNet-B1 achieved the highest F1-score with 96%, whereas EfficientNet240 had greater recall and fewer false negatives. A high-performance, computationally efficient framework for field-ready mobile or IoT devices is created by these discoveries. Combining advanced deep learning with local agricultural conditions, this work delivers a powerful, scalable, and practical technique for real-time plant species identification in intercropping farms. The findings advance smart agriculture and give farmers in small communities’ accurate agricultural technology integration.

This study presents a hybrid clustering model that integrates K-Means and HDBSCAN with embeddings generated from a transformer-based language model (BERT) to classify and cluster student queries more effectively. The approach leverages the efficiency of K-Means for initial grouping and the refinement capabilities of HDBSCAN for improving semantic granularity. Using a dataset of over 87,599 student queries and the findings shows that Silhouette Score more than doubles <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(0.3547 \rightarrow 0.7605)$</tex>, suggesting a major improvement in cluster cohesion and separation. Calinski-Harabasz Index improve from <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(1209.2181 \rightarrow 3622.5364)$</tex>, showing greater separation between clusters. Davies-Bouldin Score reduces by more than <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$65 \%(0.9592 \rightarrow 0.3278)$</tex>, indicating hybrid clusters are compact and less overlapping. These results indicate enhanced cluster cohesion and separation, supporting the viability of hybrid clustering models in educational AI systems.

This paper introduces the development and assessment of a hybrid model that integrates BERT (Bidirectional Encoder Representations from Transformers) and Naive Bayes for detecting fake news. The model utilizes BERT's deep contextual embeddings and Naive Bayes effective classification to improve detection accuracy and lower computational complexity. The paper highlights the performance of the model on the ISOT Fake News Dataset, testing the accuracy, precision, recall, and F1 score metrics, along with the efficiency of the model in the training and prediction phases. The results show that the hybrid model performs better than single models, with a high accuracy of 98.3 % and low latency inference of 0.02 s, offering a strong solution to fake news detection challenges. Lastly, the study claims the importance of integrating practical natural language processing methods to enhance misinformation management in real-world applications.

When it comes to power systems, both overall performance and efficiency are key factors in the design of the power device drivers. Using multi-level power gating logic, this study developed an active gate driver (AGD) design that is capable of driving the power transistor at different voltage levels. To lessen the amount of power that the AGD consumes, it is equipped with an effective power gating control logic. Additionally, the Four-stage Stability Control Circuit inside of the AGD makes certain that each and every power MOS device contained within the buffer arrays of the Output Stage is equally utilized. For the implementation of the AGD design, the TSMC 180-nm High-voltage (HV) Bipolar-CMOS-DMOS (BCD) process was utilized. The measurement outcomes with [Formula: see text] equal to 2[Formula: see text]nF and an operating PWM frequency of 400[Formula: see text]kHz validate the functionality of the AGD design. It dissipates an average of 533[Formula: see text]mW, which is a combination of static and dynamic power. Finally, the power gating mechanism offers a power reduction of 49.23%, as determined by a comparison of two cycles with and without power gating circuit installed.