Journal of Technology and Informatics (JoTI)

Development of Quibot Technology-Based Learning Media to Improve Student Motivation and Learning Outcomes

2026-01-06T14:47:10+07:00

This study examines the effect of Quibot, a chatbot-based learning medium, on students’ learning outcomes and learning motivation in mathematics learning at the junior high school level. The study was motivated by the limited integration of interactive digital media and its impact on students’ academic performance. A quantitative quasi-experimental design was employed involving 60 eighth-grade students of Yayasan Perguruan Sultan Iskandar Muda, divided into an experimental group using Quibot and a control group using conventional learning methods. Learning outcomes were measured using a validated multiple-choice mathematics achievement test, while learning motivation was assessed through a Likert-scale questionnaire. Data were analyzed using normality and homogeneity tests, followed by one-way ANOVA for learning outcomes and Welch’s t-test for learning motivation. The results showed a significant difference in learning outcomes between the two groups, F(1,58) = 4.65, p = 0.035, with a moderate effect size (η² = 0.074). However, no significant difference was found in learning motivation (p = 0.871). These findings indicate that Quibot effectively improves students’ academic achievement but does not significantly influence learning motivation within a short intervention period. This study highlights the role of chatbot-based learning media as an instrumental tool in supporting cognitive learning outcomes.

Implementation of The Topsis Algorithm In A Car Purchase Decision-Making System

2026-01-28T14:34:39+07:00

Private vehicles such as cars and motorcycles are crucial modes of transportation for the movement of goods and people. With technological advancements, car manufacturers offer a wide range of vehicles. Therefore, prospective buyers face challenges in selecting a vehicle that best suits their preferences and criteria. To tackle the issue, this study develops a practical decision support system (DSS) as a user-friendly tool for buyers, with theoretical contributions in the form of a more adaptive TOPSIS application and systematic analysis in car selection. This study focuses on collecting car-related data using 12 criteria, such as price, fuel consumption, safety, and design. The TOPSIS method is then normalized to ensure a fair and objective comparison between criteria. The results show the top alternative ranking, Suzuki 2002 (closeness score of 0.7089 in position 1), and the SUS test result of 85.6, indicating that the system is easy to use and capable of providing recommendations that align with user preferences. Therefore, this study highlights that the TOPSIS method can be an effective tool in supporting car purchase decision-making and making it easier for prospective buyers to choose the car that best suits their needs.

Brush-shaped Motion Gesture of UGV Using Hand Gesture Recognition

2026-01-23T13:53:19+07:00

Manual observation of corn leaf diseases in agricultural fields often faces challenges related to time, effort, and accuracy. To address these challenges, brush-shaped motion patterns, such as zig-zag and boustrophedon trajectories, provide an effective solution by enabling uniform area coverage while reducing redundant traversal, energy consumption, and sensing gaps, making them well-suited for precision agriculture applications. Building on this approach, the system utilizes the MediaPipe framework for hand landmark tracking and the K-Nearest Neighbors (KNN) algorithm to recognize six navigation commands: forward, backward, stop, turn_right, turn_left, and capture. These commands are transmitted via Wi-Fi with an average latency of 0.001964 s. To ensure navigation accuracy during pattern execution, corrections are made using rotary encoders. Gesture classification experiments on 6,000 samples achieved a maximum accuracy of 99.46% across two participants, with stable KNN performance under both indoor and outdoor lighting variations, as well as hand distances ranging from 50 cm. Furthermore, the capture gesture produced an average image acquisition latency of 0.3037 s at various UGV observation positions. In summary, these results demonstrate that integrating real-time gesture control with UGV maneuvers enables systematic field surveys for maize leaf disease monitoring and supports Sustainable Development Goal (SDG) 2 through precision agriculture technology.

Blocking Delay Effects on Microcontroller Speed and Responsiveness in Industrial IoT Devices: A Systematic Review

2026-03-20T15:36:11+07:00

This review synthesizes research on the impact of blocking delay on microcontroller speed and responsiveness in IoT devices for industrial automation. It evaluates blocking delay effects on microcontroller performance. The review benchmarks scheduling and edge computing techniques, identifies mitigation strategies, compares case study outcomes, and analyzes architectural and software factors influencing blocking delay. A systematic analysis of experimental, simulation, and co-design studies was conducted. The analysis focused on real-time scheduling, interrupt handling, network-induced latency, and edge computing integration. Key findings reveal that advanced scheduling algorithms and interrupt nesting significantly reduce blocking delays and improve task responsiveness. Edge computing and hardware optimizations also minimize network-induced latency and enhance local processing capabilities. Multiple sources of blocking delay, including resource contention and network overload, are mitigated through adaptive scheduling and hardware-assisted mechanisms. Real-world case studies confirm substantial latency reductions and improved control performance in industrial IoT contexts. These findings underscore the interplay of software and hardware factors in shaping microcontroller responsiveness. The review highlights the necessity for scalable, integrated solutions that address dynamic industrial environments. It informs the design of more responsive and efficient microcontroller-based IoT systems for industrial automation.

Comparative Performance of Machine Learning Algorithms for Diabetes Prediction

2026-03-26T12:52:46+07:00

Early detection of diabetes mellitus is crucial to prevent severe complications. This study evaluates three machine learning algorithms for diabetes prediction using a quantitative comparative experimental design. The algorithms are k-Nearest Neighbors (k-NN), Support Vector Machine (SVM), and Random Forest. These methods were chosen to compare distinct learning paradigms. k-NN is distance-based, SVM is margin-based, and Random Forest is an ensemble method. The goal is to find the optimal model for clinical use. The Pima Indians Diabetes dataset was used. It includes 390 patients and 15 clinical features. Performance was measured by accuracy, precision, recall, and F1-score. Random Forest had the highest accuracy (89.7%) and F1-score, providing the most balanced classification. SVM followed with 84.6%, and k-NN achieved 76.9%. Although k-NN had the highest recall (0.750), its precision was low (0.375), showing a high false-positive rate. Feature importance analysis pointed to blood glucose levels as the most significant predictor, which matches clinical knowledge. In summary, ensemble techniques like Random Forest offer the most reliable results. This highlights the importance of selecting the right algorithm for early diabetes detection in clinical applications.