Rancang Bangun Sistem Traffic Monitoring AGV Berbasis Integrasi PLC-ESP32 melalui Visualisasi Waypoint secara Real-Time Menggunakan MQTT

Abstract

Efisiensi pemantauan posisi dan status Automated Guided Vehicle (AGV) merupakan faktor krusial bagi kelancaran produksi manufaktur. Penelitian ini mengatasi keterbatasan visibilitas dengan membangun sistem traffic monitoring berbasis IoT untuk satu unit AGV sebagai prototipe pada jalur Door Lock PT Aisin Indonesia. Sistem ini mengintegrasikan PLC Mitsubishi FX5U dan ESP32 dengan menyajikan data lokasi secara real-time. Metode mencakup perancangan perangkat keras dengan relay sebagai isolator, protokol MQTT, serta dashboard kustom berbasis Laravel. Akurasi visualisasi dioptimalkan melalui 63 titik waypoint dengan algoritma Linear Interpolation (LERP) guna menghasilkan pergerakan ikon halus. Selain itu, logika safe gap management divalidasi sebagai konsep proteksi visual berbasis simulasi multi-AGV, belum sebagai proteksi fisik final pada PLC untuk memitigasi risiko tabrakan melalui notifikasi otomatis. Hasil pengujian menunjukkan keandalan sistem dengan rata-rata latensi transmisi sender-to-receiver sebesar 292,34 ms dengan standar deviasi sampel sebesar 27,02 ms dan tingkat kesesuaian skala koordinat digital dengan toleransi pergeseran di bawah 3 cm terhadap denah teknis asli berdasarkan simulasi dan analisis pemetaan 63 waypoint. Validasi status operasional menggunakan confusion matrix menghasilkan akurasi 93,7%, yang menunjukkan bahwa sistem sangat andal dalam mengenali status AGV, meskipun masih ditemukan keterbatasan klasifikasi berupa false idle dan delayed stop pada titik transisi waypoint. Implementasi ini berpotensi meningkatkan efisiensi operasional dengan memangkas waktu deteksi sebesar 95,00–98,33% dengan rata-rata persentase mencapai 96,19%. Secara keseluruhan, sistem ini berhasil meningkatkan kesadaran situasional dan menurunkan nilai Mean Time to Detect (MTTD) pada Divisi Equipment Development (EQDEV).

The efficiency of monitoring the position and status of Automated Guided Vehicles (AGVs) is a crucial factor for smooth manufacturing production. This study addresses visibility limitations by developing an IoT-based traffic monitoring system for a single AGV unit as a prototype on the Door Lock line at PT Aisin Indonesia. The system integrates a Mitsubishi FX5U PLC and an ESP32 to present near real-time location data. The methodology includes hardware design utilizing relays as galvanic isolators, the MQTT protocol, and a custom Laravel-based dashboard. Visualization accuracy is optimized through 63 waypoint utilizing the Linear Interpolation (LERP) algorithm to produce smooth icon movements. In addition, the safe gap management logic is validated as a simulation-based multi-AGV visual protection concept—rather than a final physical protection system on the PLC—to mitigate collision risks through automatic notifications. The test results demonstrate system reliability, yielding an average sender-to-receiver transmission latency of 292.34 ms with a sample standard deviation of 27.02 ms, and digital coordinate scale consistency with a geometric shift tolerance under 3 cm against the original technical blueprint based on the simulation and mapping analysis of the 63 waypoints. Operational status validation using a confusion matrix achieved an accuracy of 93.7%, indicating that the system is highly reliable in recognizing AGV statuses, despite minor classification limitations such as false idle and delayed stop anomalies at certain waypoint transition points. This implementation offers an estimated potential reduction in detection time of 95.00–98.33%, with an average reduction of 96.19%. Overall, this system successfully improves situational awareness and decreases the Mean Time to Detect (MTTD) in the Equipment Development (EQDEV) Division.