Claim Missing Document
Check
Articles

Found 22 Documents
Search

Pemetaan Dan Pengukuran Pemanfaatan TIK Menggunakan Metode Echosystem (Studi Kasus STT-Payakumbuh) Noviardi Noviardi; Dilson Dilson; Ranti Irsa
Elkawnie Vol. 2 No. 1 (2016)
Publisher : Faculty of Science and Technology Universitas Islam Negeri Ar-Raniry

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.22373/ekw.v2i1.2519

Abstract

The rapid growth of ICT (Information and CommunicationTechnology), which changed the whole aspect of human life, such as the way people think, behave, communicate, work and support themselves. This change had a positive impact on the improvement of the efficiency of information and communication needs. So that the use of IT in all areas of life provides a significant advantage. Utilization of Information and Communication Technology (ICT) in the institution of Higher Education requires planning and management are mature, both in infrastructure, human resources and in terms of users, so that the use of ICT / ICT not only to improve the effectiveness, efficiency and productivity, but also can improve competitiveness (competitive adventages) in the era of globalization ini.Metode used in this study is a Echosystem consisting of three (3) stages of development that the ICT development by UNSECO, iT development, and the development model Zen Framework Smart Campus (TeSCA). It is hoped this method can be a strategic step in the utilization of and basis for the advancement of the ICT development of STT-Payakumbuh campus in the future
Energy-Aware Adaptive TinyML pada IoT Bertenaga Surya untuk Optimalisasi Edge AI Pertanian Presisi Noviardi; Rosda Syelly; Habilillah
Technologica Vol. 5 No. 2 (2026): Technologica
Publisher : Green Engineering Society

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55043/technologica.v5i2.623

Abstract

Penerapan Tiny Machine Learning (TinyML) pada perangkat IoT bertenaga surya sering terhambat oleh fluktuasi energi dan lonjakan arus aktuator mekanis. Sistem Edge AI konvensional dengan inferensi statis sangat rentan memicu kegagalan daya total (brown-out). Penelitian ini mengusulkan sebuah arsitektur siber-fisik lintas-lapisan yang memungkinkan node IoT mengadaptasi beban kerja kognitif dan fisik secara otonom berdasarkan kesadaran kapasitas energi internal (State of Charge/SoC). Diimplementasikan pada ESP32, sistem ini mengorkestrasi jaringan saraf multi-tingkat (Model Gold, Silver, dan Bronze) yang dikuantisasi ke presisi INT8 secara lossless. Sistem secara mandiri melakukan hot-swapping komputasi sesuai ketersediaan daya, seraya memodulasi beban aktuator irigasi tetes via sinyal Pulse Width Modulation (PWM). Observasi lapangan membuktikan efektivitas sistem; manajemen daya menunjukkan kinerja baik dengan dominasi Model Silver (97,89%) dan eksploitasi surplus daya via Model Gold (0,03%). Keseimbangan komputasi ini memastikan SoC tidak pernah menyentuh defisit kritis (aktivasi Bronze 0,00%), sehingga tidak ditemukan brown-out selama pengujian. Didukung mitigasi perangkat keras dioda penyearah, efisiensi konversi daya sistem keseluruhan mencapai 96,06%, mendukung tercapainya Energy-Neutral Operation (ENO) selama observasi. Secara agronomis, intervensi hard-cutoff PWM saat batas kelembaban tanah 80% terpenuhi menghasilkan Water Use Efficiency (WUE) sebesar 0,3141 %RH/Liter tanpa memicu limpasan permukaan. Kerangka ini menawarkan cetak biru komputasi edge adaptif yang andal untuk ekosistem pertanian presisi off-grid