Stephano C. W. Ngangi
Unknown Affiliation

Published : 2 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 2 Documents
Search

SISTEM DETEKSI HAMA PADA TANAMAN JAGUNG (ZEA MAYS) BERBASIS KECERDASAN BUATAN DAN INTERNET OF THINGS (IOT) Tenda, Edwin; Stephano C. W. Ngangi; Christian A. J. Soewoeh; Ketaren, Eliasta
Jurnal TIMES Vol 14 No 2 (2025): Jurnal TIMES
Publisher : STMIK TIME

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Deteksi hama jagung secara dini pada perangkat edge sangat penting untuk pertanian presisi. Penelitian ini bertujuan untuk mengembangkan dan mengimplementasikan sistem klasifikasi hama jagung real-time yang ringan pada Raspberry Pi 4. Metode penelitian melibatkan penggunaan dataset citra dari Kaggle yang difilter menjadi tiga kelas target: ulat grayak, penggerek batang, dan daun sehat. Model Convolutional Neural Network (CNN) ringan, YOLOv8n-cls, dilatih selama 50 epoch menggunakan metode validasi hold-out (80/20) dan augmentasi data dinamis. Model best.pt yang telah dilatih kemudian dikonversi ke format ONNX untuk optimalisasi inferensi CPU. Hasil evaluasi menunjukkan performa akurasi keseluruhan yang sangat tinggi, mencapai ~99% (Top-1) dan 100% (Top-5). Meskipun demikian, analisis confusion matrix mengungkap adanya ketidakseimbangan data (dataset imbalance) yang ekstrem. Model menunjukkan recall 100% pada kelas dominan (ulat grayak, 119 sampel uji), namun performa pada kelas minoritas (daun sehat 85.7% dan penggerek batang 100%) tidak signifikan secara statistik karena jumlah sampel uji yang sangat sedikit (masing-masing 7 dan 1). Sistem ini berhasil diimplementasikan pada Raspberry Pi 4 menggunakan OpenCV dan ONNX Runtime, mencapai latensi inferensi yang rendah (~40-50 ms). Disimpulkan bahwa meskipun model sangat akurat dan cepat dalam mendeteksi ulat grayak, akurasi keseluruhannya bersifat miring (biased) dan diperlukan penambahan data signifikan untuk kelas minoritas agar sistem dapat diandalkan secara penuh di lapangan.
Perancangan dan Evaluasi Usability UI/UX Aplikasi Manajemen Produksi Budidaya Sarang Burung Walet Menggunakan Metode Design Thinking (Studi Kasus: CV Mega Walet Sejahtera) Ivanka A. J. Pasanda; Mans L. Mananohas; Stephano C. W. Ngangi
INTECH Vol. 7 No. 1 (2026): INTECH (Informatika Dan Teknologi)
Publisher : Informatics Study Program, Faculty of Engineering and Computers, Baturaja University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.54895/intech.v7i1.3436

Abstract

CV Mega Walet Sejahtera still manages the recording and production management process of edible bird’s nest cultivation manually using Microsoft Excel and separate communication media, causing delays in reporting, difficulties in data recapitulation, and inefficient production monitoring and evaluation processes. This study aims to design and evaluate the usability of a UI/UX prototype for an edible bird’s nest production management application using the Design Thinking method. The research applied a mixed methods approach through interviews, observations, usability testing, and the System Usability Scale (SUS) questionnaire. The design process was carried out through the stages of empathize, define, ideate, prototype, and testing. The application prototype was designed using Figma and tested using the Maze platform involving seven respondents consisting of the business owner and field workers. The results showed that the application prototype achieved a very good level of usability based on Nielsen’s five usability attributes, namely learnability, efficiency, memorability, errors, and satisfaction. The SUS testing produced an average score of 87.7, which falls into the best imaginable category and is considered acceptable. In addition, the MAUS testing results obtained scores of 93 for the owner and 89 for the workers. These findings indicate that the Design Thinking method successfully produced a UI/UX design that meets user needs and supports more effective and user-friendly production recording, reporting, and monitoring processes.