INTI Institutional Repository
Not a member yet
2178 research outputs found
Sort by
Developing Low-Cost LoRaWAN Internet of Things Devices for Water Resources Monitoring in Bali
No full textIn this study, we developed a solar-powered prototype using an ESP-32 MCU, commercially
available sensors, and a LoRaWan communication module. The components cost less than $30
USD. The prototype has been running on solar power for two months in room conditions, repeating
the sleep-wake cycle and transmitting sensor data - temperature, battery %, light color, and
accelerometric data - every ten minutes over LoRaWAN to a cloud data storage. While the data only reflect room settings, and not real environmental data, the operating record demonstrates
steady behavior, power autonomy, and data transfer, which is a necessity for IoT devices that monitor water supplies in the field. In the future, the developed devices will be used in Bali,Indonesia, to monitor the hydrological status during an impending water crisis
Data-Driven Analysis of Computer-Based Testing to Advance Machinist Performance
No full textThe rapid advancement of technology has transformed the education sector, offerings new avenues
for data-driven teaching and learning innovations. This study investigates the integration of
Augmented Reality (AR) technology in developing an interactive learning media application for
scout password recognition, with a focus on analyzing learner interaction data to evaluate its
effectiveness. The application utilizes marker-based tracking to overlay digital content in the real
world, creating an immersive environment that enhances comprehension and retention. The study
employs the Prototype Method to ensure user-centric design, supported by stakeholder feedback
throughout iterative development. Unified Modeling Language (UML) tools, such as Use Case
and Activity Diagrams, were utilized to model system functionality. Key features of the application
include interactive 3D models, gamification elements, and progress tracking, with user interaction
data analyzed to assess engagement and learning outcomes. System functionality was evaluated
using the Blackbox testing method, and user performance data was analyzed to identify patterns
in engagement, motivation, and understanding of scout passwords. Results reveal a significant
improvement in learner outcomes compared to traditional teaching methods, with data analysis
highlighting areas of particular effectiveness, such as the use of gamification to sustain learner
interest. This research not only underscores the potential of AR in transforming niche educational
contexts but also emphasizes the importance of analyzing interaction and performance data to
refine educational tools. Future development recommendations include incorporating AI-powered
personalized learning features and expanding the application to cover additional scouting skills,
paving the way for broader adoption of AR technology in education
Evaluation and Comparative Analysis of Feature Extraction Methods on Image Data to increase the Accuracy of Classification Algorithms
No full textManual selection of fresh fruit has been identified as a significant challenge for the agricultural
sector due to its time-consuming nature and potential for inconsistent categorization. This process
requires human labour to visually inspect and sort fruits, leading to variability and inefficiencies
in the sorting process. This research proposes a low-cost alternative using intelligent fruit selection
systems based on computer vision techniques to address these issues. These systems aim to
automate the process of fruit selection, improving efficiency and consistency in categorizing fruits
such as apples, bananas, and oranges. A critical step in developing such intelligent systems is the
feature extraction process. Feature extraction is essential in classification, especially for data
sources in the form of images. It involves identifying and isolating relevant information from the
images that classification algorithms can use to distinguish between different fruit categories. If
the feature extraction process fails to capture the correct information, the performance or accuracy
of the classification algorithm will be negatively impacted. This research compares three different
methods for extracting features from fruit images to determine which method yields the highest
accuracy for fruit classification. The feature extraction methods evaluated were Grayscale Pixel
Values, Mean Pixel Value of Channels, and Extracting Edge Features. The classification algorithm
used in this research is the Convolutional Neural Network (CNN) algorithm. CNNs are well-suited
for image classification tasks due to their ability to learn hierarchical feature representations from
the input images automatically. By comparing the performance of the CNN classifier using the
three different feature extraction methods, this research aims to identify the method that provides
the highest level of accuracy in classifying fruit images. By conducting this comparative analysis,
the research provides insights into the most effective feature extraction techniques for improving
the performance of computer vision-based fruit selection systems, ultimately contributing to more
efficient and consistent fruit categorization in the agricultural sector. The result revealed that the
Grayscale achieved the highest validation accuracy (99.94%) and the lowest validation loss
(0.44%)
Machine Learning Models for Classification of Anemia from CBC Results: Random Forest, SVM, and Logistic Regression
No full textIn an effort to increase diagnostic efficiency and accuracy, this work investigates the
application of machine learning models Random Forest, SVM, and Logistic Regression for the
categorization of anemia. Hematocrit and hemoglobin levels were included in the dataset, which
was divided into training and testing sets. Using CatBoost, Random Forest outperformed SVM
(82.1%) and Logistic Regression (75.1%) with the greatest accuracy (99.2%). SVM and Logistic
Regression work well with simpler data, while Random Forest performs best with intricate medical
datasets, which makes it perfect for applications involving the detection of anemia
Potential of Tabas Stone Waste as Additional Material of Concrete for Coastal Protection Structures
No full textThe coast has natural protection, but if this natural protection is damaged, the coast can be
protected with coastal structures. Concrete is one of the main materials for coastal protection
structures such as breakwaters, jetties, groins and revetments. Concrete used in coastal
environments must have high strength performance to face the challenges of corrosive seawater,
high humidity, and extreme temperature changes. Tabas stone is a Basaltic Scoria stone resulting
from the eruption of Mount Agung which is used by the people of Bali as an ornament in Balinese
buildings. Tabas stone pieces that do not match the size are discarded and become waste. In this
study, tabas stone waste was used as an additional material for fine aggregate of 0%, 10%, 20%
with a concrete design compressive strength of 42 MPa. Cylindrical samples were produced then
soaked in the sea and at the river mouth. Furthermore, the samples were tested to be compared
with concrete samples with curing in standard water conditions at the age of 28 days. The test
results showed that the effect of seawater and brackish water immersion caused a significant
decrease in the compressive strength of the concrete. The addition of the percentage of tabas stone
also caused a decrease in the compressive strength of the concrete. Thus, tabas stone have small
potential to be used as an additional material for concrete filler for coastal building construction
Recognize Hate Speech On Twitter Using Machine Learning
No full textConvolutional Neural Network (CNN) is a frequent-deep learning algorithm that is powerful
in classifying image and text data, the system analyses individual tweets in order to determine
if it contains hate speech. The occurrence of offensive speech in online forums poses significant
challenges to maintaining a safe and inclusive digital environment. This study addresses these
challenges by developing a hate speech recognition system ML methods, specifically CNN
algorithms aimed primarily at analysing hate speech in tweets, attempting to increased resource
efficiency and accuracy, its system analyses textual content in the tweet and produces and
indicates whether it contains hate speech and determines the percentage of intolerance speech
present in the tweet. The results of this study highlight the power of CNN-based strategies in
preventing cyberbullying and promoting healthy digital discourse
Factors Influencing Industrial Waste Applying Information Technology and Managing Information Systems Towards Minimizing Waste Management
No full textThis study explores the surging of demand for manufactured products with the increasing of
world’s population. While the manufacturing sector is essential in meeting this demand, it also
faces the significant challenge of reducing the environmental impact of industrial waste.
Manufacturing industrial waste primarily arises from by-products, over-extraction of natural
resources, and inefficient production processes. Poor management of manufacturing processes
would lead to adverse social impacts to human health, natural resources depletion, ecosystem
destruction, and contribute to global warming and climate change. Additionally, improper waste
disposal can result in financial losses and legal penalties for non-compliance with environmental
regulations. Hence, it is imperative for the manufacturing sector to leverage emerging technologies
and management strategies to mitigate these challenges. Therefore, this study used a quantitative
approach to analyse primary data collected from the survey questionnaire to examine the
relationship between independent and dependent variables. Combination of financial, technical,
social, and governmental factors addressed in this project underscores the multifaceted approach
required for effective waste managemen
Analysis of New Student Admission Application for Bina Warga High School in Palembang
No full textThe manual admission process at Bina Warga 2 High School Palembang poses significant
challenges, including inefficiencies in registration, limited accessibility for prospective students,
and increased administrative workload. To address these issues, this study analyses the
requirements for an online student admission system designed to streamline registration, facilitate
efficient report generation, and serve as a promotional platform for the school. By adopting the
Prototyping method for system development, the process ensures active collaboration between
developers and stakeholders, enabling iterative refinement based on user feedback. The proposed
system eliminates the need for prospective students to visit the school physically, thereby
enhancing accessibility and scalability. Furthermore, it automates administrative tasks, reducing
manual effort and improving overall efficiency. This analysis highlights the potential of an online
admission system to transform the registration process, expand the school’s outreach, and establish
a robust digital presence for future growth
Enhancing Travel Recommendations Through Attraction Preference Standardization
No full textThe subsequent paper describes a typical travel recommendation system that comprises
collaborative filtering, content-based filtering, and sentiment analysis in its design. The
proposed system enhances the problems of conventional methods by using operation
preferences to standardize attraction and adopting sentiments obtained from the rating. The
following are the steps of the implementation of the study; data collection, data pre-processing,
data modelling and the last is the development of web application. Actual analysis proves that
there was a general enhancement in the precision of the recommendation and also the
satisfaction level of the customers
Implementation of Health Monitoring System for Patients using Machine Learning Algorithms
No full textTo enhance monitoring and forecasting skills, we investigate in this research study the inclusion
of cutting-edge technology in the industrial and healthcare domains. We created a machinelearning-
based solution for the wellness program industry that uses Internet Of Medical Things
(IoMT) devices to forecast cardiovascular risk. Our model outperformed previous approaches
in diagnosing cardiovascular disease (CVD) with higher accuracy, recall, and F1-score. It did
this by using a fuzzy logic classifier for illness prediction and a random forest for feature
selection. Additionally, to enhance overall equipment effectiveness (OEE), lower electricity
costs, and decrease unplanned downtime in manufacturing settings, we created a real-time
system leveraging smart systems and machine learning. During testing on a manufacturing
blender, this device tracked operational phases and load-balancing conditions well. We
employed the Decision Tree Algorithm to train and assess a model that produced a perfection
of 66.66%