IPB University Logo

SCIENTIFIC REPOSITORY

IPB University Scientific Repository collects, disseminates, and provides persistent and reliable access to the research and scholarship of faculty, staff, and students at IPB University

AI Repository
 
Building and Categories


      View Item 
      •   IPB Repository
      • Final Assignments
      • Undergraduate Final Assignments
      • UF - Vocational School
      • UF - Computer Engineering Tehcnology
      • View Item
      •   IPB Repository
      • Final Assignments
      • Undergraduate Final Assignments
      • UF - Vocational School
      • UF - Computer Engineering Tehcnology
      • View Item
      JavaScript is disabled for your browser. Some features of this site may not work without it.

      Implementasi Logika Dwell Time pada ESP32-CAM untuk Reduksi Alarm Palsu pada Sistem Keamanan Gudang Berbasis Edge-Cloud

      Thumbnail
      View/Open
      Cover (379.3Kb)
      Fulltext (1.697Mb)
      Lampiran (248.5Kb)
      Date
      2026
      Author
      Fiihudan, Elegan Basyaton
      Fathonah, Lathifunnisa
      Metadata
      Show full item record
      Abstract
      Keamanan gudang pasif untuk aset bernilai tinggi sering terhambat oleh sistem pemantauan konvensional yang bersifat reaktif serta tingginya tingkat alarm palsu dari sensor gerak Passive Infrared (PIR). Penelitian ini bertujuan mengembangkan prototipe sistem keamanan hibrida Edge-Cloud yang proaktif dan efisien untuk memvalidasi ancaman. Logika dwell time 12 detik diterapkan langsung pada firmware ESP32-CAM sebagai lapisan penyaring pertama di sisi Edge sehingga aktivitas transit normal ditahan dan hanya perilaku loitering yang dilaporkan. Gambar yang lolos penyaringan diunggah ke Supabase, kemudian divalidasi model YOLOv8n pada server DigitalOcean sebagai lapisan kedua, dan hasilnya disampaikan melalui aplikasi web Next.js serta Telegram Bot API. Hasil pengujian menunjukkan sistem menahan seluruh pemicu pada skenario transit normal (100%), sedangkan validasi visual mencapai akurasi 100% pada kondisi siang dan sore, tetapi 0% pada malam hari akibat keterbatasan pencahayaan LED flash bawaan. Analisis latensi menunjukkan proses unggah dari perangkat Edge dengan rata-rata 13,1 detik menjadi hambatan utama sistem, jauh melampaui waktu inferensi YOLOv8n yang hanya 295 ms.
       
      Passive warehouse security for high-value assets is often hindered by conventional monitoring systems that are reactive in nature and by the high falsen alarm rate of Passive Infrared (PIR) motion sensors. This study aims to develop a proactive and efficient hybrid Edge-Cloud security prototype for validating threats. A 12-second dwell time logic was implemented directly in the ESP32-CAM firmware as the first filtering layer on the Edge side, so that normal transit activity is withheld and only suspicious loitering behaviour is reported. Images passing this filter are uploaded to Supabase, validated by a YOLOv8n model on a DigitalOcean server as the second layer, and delivered to the user through a Next.js web application and the Telegram Bot API. Testing showed that the system withheld all triggers in the normal transit scenario (100%), while visual validation achieved 100% accuracy in daytime and afternoon conditions but 0% at night due to the limited illumination of the onboard LED flash. Latency analysis showed that the upload process the Edge device, averaging 13.1 seconds, is the main bottleneck, far exceeding the YOLOv8n inference time of only 295 ms.
       
      URI
      http://repository.ipb.ac.id/handle/123456789/179567
      Collections
      • UF - Computer Engineering Tehcnology [259]

      Copyright © 2020 Library of IPB University
      All rights reserved
      Contact Us | Send Feedback
      Indonesia DSpace Group 
      IPB University Scientific Repository
      UIN Syarif Hidayatullah Institutional Repository
      Universitas Jember Digital Repository
        

       

      Browse

      All of IPB RepositoryCollectionsBy Issue DateAuthorsTitlesSubjectsThis CollectionBy Issue DateAuthorsTitlesSubjects

      My Account

      Login

      Application

      google store

      Copyright © 2020 Library of IPB University
      All rights reserved
      Contact Us | Send Feedback
      Indonesia DSpace Group 
      IPB University Scientific Repository
      UIN Syarif Hidayatullah Institutional Repository
      Universitas Jember Digital Repository