Jurnal Teknik Informatika C.I.T Medicom

Analysis of Household Energy Consumption Patterns Using K-Means Clustering and Explainable Data Mining

2026-06-20T09:18:40+00:00

The increasing demand for household energy has created significant challenges for energy sustainability, resource management, and the development of effective energy efficiency strategies, thereby necessitating advanced analytical approaches to better understand residential consumption behavior. This study aims to analyze household energy consumption patterns using K-Means clustering and explainable data mining techniques. Household energy consumption data were collected from residential users and subjected to preprocessing procedures, including data cleaning, missing value handling, feature selection, and normalization to ensure data quality and analytical reliability. The K-Means clustering algorithm was then applied to identify homogeneous groups of households based on their energy consumption characteristics, while explainable data mining techniques were employed to interpret cluster profiles and determine the factors influencing cluster membership. The results revealed the existence of three distinct household energy consumption groups, namely low-, moderate-, and high-consumption households, each exhibiting significantly different consumption behaviors, appliance ownership levels, and energy expenditure patterns. Further analysis showed that household size, appliance ownership, and peak electricity usage were the most influential factors differentiating the clusters. These findings demonstrate that the integration of K-Means clustering and explainable data mining provides an effective and interpretable framework for understanding household energy consumption behavior. The proposed approach offers valuable insights for utility companies, policymakers, and consumers by supporting targeted energy efficiency programs, demand-side management initiatives, and evidence-based energy policy development aimed at promoting sustainable household energy consumption.

Real-time human detection on FPV drones using YOLOv11 and ESP-NOW

2026-06-13T15:23:10+00:00

Conventional aerial surveillance systems still rely heavily on human operators, which may lead to visual fatigue, limited monitoring coverage, and delayed responses during security patrol operations. This study proposes a real-time human detection system for FPV drone surveillance using the YOLOv11 object detection model integrated with ESP-NOW wireless communication. The proposed system incorporates temporal validation and human-in-the-loop confirmation to improve detection reliability and maintain operator control during response activation. Experimental evaluations were conducted under morning, afternoon, and evening conditions. The proposed system achieved average confidence values of 81.25%, 78.38%, and 79.88%, with detection success rates of 71.13%, 75.94%, and 78.03%, respectively. Furthermore, the ESP-NOW communication subsystem successfully transmitted activation signals with delays ranging from 7 ms to 53 ms and maintained stable communication over distances up to 300 m. The main contribution of this research lies in the integration of YOLOv11, temporal validation, human-in-the-loop confirmation, and ESP-NOW communication into a single UAV surveillance framework, enabling reliable real-time human detection while preserving human supervision in operational decision-making.

Design and implementation of a weapon storage access control system based on hand gesture recognition and face recognition on Raspberry Pi 5

2026-06-16T01:19:55+00:00

This study presents a multimodal biometric access control system for weapon storage facilities, integrating hand gesture recognition and face recognition through a sequential fusion architecture on Raspberry Pi 5. The sequential design activates face verification only after correct gesture authentication, optimizing computational efficiency on edge hardware while establishing a dual-layer security barrier. The gesture module combines MediaPipe Hands landmark extraction with LSTM-based temporal classification, achieving near-perfect accuracy across four gesture classes. The face module employs dlib's ResNet-34 for 128-dimensional embedding comparison, with an empirically recalibrated Euclidean distance threshold of 0.34 to eliminate false acceptance risks identified during intrusion testing. Evaluation under controlled conditions yielded 0% False Reject Rate and 0% False Accept Rate across 60 trials, with reliable GPIO-controlled solenoid actuation. Results demonstrate that sequential fusion of behavioral and physiological biometrics on a single edge device provides a viable security solution for high-risk access control applications.

Implementation of BLAKE3 hashing for accelerating digital evidence integrity verification in forensic investigations

2026-06-18T01:22:50+00:00

The evolution of cybersecurity threats demands rapid and legally accountable investigation responses. A crucial principle in digital forensics is maintaining data integrity to ensure the validity of the chain of custody in court using cryptographic hash functions. However, the increasing volume of storage media presents significant technical challenges. Conventional algorithms like SHA-256 process data sequentially, causing hash verification on massive forensic images to take hours. This study aims to evaluate the BLAKE3 algorithm as an accelerator in the digital evidence integrity verification process. The evaluation was conducted using a comparative experimental method between MD5, SHA-256, and BLAKE3 by varying processor core allocations and simulated file sizes up to 50 GB. The test results demonstrated that parallel processing in BLAKE3 significantly reduces execution time. In the 50 GB file test utilizing 8 threads, BLAKE3 achieved a throughput of 5000 MB/s and completed verification in just 10.0 seconds, vastly outperforming SHA-256 which required 142.8 seconds. The application of BLAKE3 proved to provide security equivalent to SHA-256 while accelerating the verification process, thereby supporting more efficient courtroom proceedings without violating legal integrity standards.

Development of an Explainable Expert System for Smart Factory Readiness Assessment in Manufacturing Industries

2026-06-22T13:35:48+00:00

Smart Factory adoption has become a critical strategy for manufacturing industries seeking to improve productivity, operational flexibility, and global competitiveness in the era of Industry 4.0. However, many organizations still lack a systematic, reliable, and transparent approach to evaluating their readiness for Smart Factory implementation. This study aims to develop an Expert System for determining industry readiness for Smart Factories by integrating expert knowledge and Explainable Artificial Intelligence (XAI). Expert knowledge was acquired through interviews with Industry 4.0 specialists, manufacturing practitioners, and automation experts, as well as an extensive literature review to identify readiness criteria related to technology, organization, human resources, processes, and financial capability. To enhance transparency and user trust, Explainable AI techniques were incorporated to provide interpretable explanations and feature contribution analyses for readiness recommendations. The system was validated through expert evaluation and case studies involving manufacturing organizations. The results indicate that the proposed system successfully classified organizations into five readiness levels and generated clear, understandable explanations for each recommendation. Validation findings demonstrated a high level of agreement between system outputs and expert assessments, confirming the reliability and practical applicability of the proposed approach. Furthermore, feature contribution analysis revealed that automation level, workforce digital skills, and IoT infrastructure were the most influential determinants of Smart Factory readiness.

Decision Support System for Determining Cyber Risk Mitigation Priorities in Higher Education Using the Fuzzy TOPSIS Method

2026-06-18T13:51:58+00:00

The increasing frequency and sophistication of cyber threats have made higher education institutions attractive targets for cyberattacks, posing significant risks to information assets, academic operations, and institutional reputation. Universities rely heavily on digital technologies, including academic information systems, e-learning platforms, cloud services, and research databases, making effective cybersecurity risk management essential. However, limited cybersecurity resources often prevent institutions from addressing all potential threats simultaneously, highlighting the need for a systematic approach to prioritizing cyber risk mitigation efforts. This study aims to develop a Decision Support System (DSS) for determining cyber risk mitigation priorities in higher education institutions using the Fuzzy Technique for Order Preference by Similarity to Ideal Solution (Fuzzy TOPSIS) method. Six evaluation criteria were considered, namely probability of occurrence, financial impact, operational impact, reputation damage, data sensitivity, and recovery complexity. Expert assessments were expressed using linguistic variables and converted into Triangular Fuzzy Numbers (TFNs) to accommodate uncertainty in the decision-making process. The Fuzzy TOPSIS method was then applied to evaluate and rank cyber risks according to their mitigation priorities. The results demonstrated that the proposed DSS successfully generated a prioritized ranking of cyber risks, with ransomware and data breach risks receiving the highest mitigation priorities due to their substantial impacts on university operations, financial resources, and information security. The findings suggest that the developed DSS effectively supports cybersecurity decision-making by handling uncertainty in expert assessments and providing systematic recommendations for cyber risk mitigation. Consequently, the proposed framework can assist higher education institutions in allocating cybersecurity resources more efficiently and enhancing their overall cybersecurity resilience.

Expert System for Identification of Digital Transformation Maturity Level in Secondary Schools Using Forward Chaining

2026-06-24T08:59:47+00:00

The rapid growth of digital transformation in education has significantly influenced secondary schools through the adoption of e-learning platforms, digital administration systems, and smart classroom technologies, making digital maturity an essential aspect of modern educational development. However, many schools still lack a structured and standardized approach to measure their level of digital transformation maturity, resulting in assessments that are often subjective and inconsistent. This study aims to develop an expert system for identifying the digital transformation maturity level of secondary schools. The system is built using a rule-based approach supported by a knowledge base derived from expert interviews and relevant literature, including established digital maturity frameworks. Data collection is conducted through structured questionnaires that capture key indicators such as ICT infrastructure, teacher digital literacy, digital learning adoption, and institutional policy support. The inference mechanism employed in the system is Forward Chaining, which processes input facts and applies IF–THEN rules to generate logical conclusions. The system is capable of classifying schools into predefined maturity levels, ranging from initial to advanced stages of digital transformation. The results indicate that the expert system can effectively evaluate and categorize digital maturity levels in a systematic and consistent manner. In conclusion, the proposed system provides a reliable decision-support tool that assists school administrators and policymakers in assessing and improving digital transformation readiness in secondary education institutions.

Machine Learning Integration in DEA Models: Current Developments and Future Challenges

2026-06-19T14:07:08+00:00

The increasing availability of large and complex datasets has created new opportunities for enhancing Data Envelopment Analysis (DEA) through the integration of Machine Learning (ML) techniques. This study reviews current developments in the integration of ML and DEA models and identifies key challenges, trends, and future research opportunities. A systematic literature review was conducted by examining recent studies that combine DEA with various machine learning algorithms across multiple application domains, including healthcare, banking and finance, manufacturing, supply chain management, energy, agriculture, and higher education. The findings indicate that Artificial Neural Networks (ANN), Support Vector Machines (SVM), Random Forests, Gradient Boosting methods, and Deep Learning models are among the most frequently employed techniques in DEA-ML frameworks. Despite these advantages, several challenges remain, including data quality issues, model interpretability, computational complexity, limited generalizability, and the lack of standardized integration frameworks. The review concludes that the integration of ML and DEA offers substantial potential for advancing efficiency analysis and organizational performance evaluation. Future research should focus on developing explainable artificial intelligence (XAI) solutions, real-time efficiency analytics, federated learning approaches, and standardized hybrid DEA-ML frameworks to improve transparency, scalability, and practical applicability across diverse operational environments.