Development of Tropical Grape Greenhouse Monitoring and Nutrient Delivery Scheduling using MQTT Protocol and IoT

Supriyanto, Supriyanto
Solahudin, Mohamad
Sucahyo, Lilis
Aryawan, Putu Oki Wiradita
Widodo, Slamet
Susanto, Slamet

Teknotan: Jurnal Industri Teknologi Pertanian
Vol 20, No 2 (2026): TEKNOTAN, Agustus 2026 (In Progress)

Grape production in Indonesia is limited due to environmental conditions and other factors. However, grape consumption in Indonesia has become popular over the last decade. Nutrient and water delivery are scheduled based on the growth stages of the grape, which consist of vegetative and generative stages. Those dynamic schedules should be assisted using automatic monitoring and controlled using the message queuing telemetry transport (MQTT) and Internet of Things protocol. However, the performance of monitoring and control should be evaluated. This study aimed to evaluate the performance of Tropical Grape Greenhouse Monitoring and Nutrient Delivery Scheduling using the MQTT protocol and IoT technique. The research stages include system analysis, design, implementation, and testing. The outcomes of this study consist of a monitoring system and a control system. The monitoring system is equipped with a main control unit (MCU) based on an ESP32, along with sensors for air temperature and humidity (DHT21), light intensity (lux sensor), water temperature (DS18B20), and soil moisture. Data is transmitted via an Aedes-based MQTT broker installed on the Node-RED platform on a virtual private server (VPS) integrated with a MySQL database. The monitoring and control system is implemented in the grape greenhouse at the Agribusiness and Technology Park (ATP) at IPB University. Specifically, the control system is designed and implemented in the grape greenhouse to activate solid-state relays (SSRs) and contactors to operate water pumps and solenoid valves, all managed through a dashboard interface. System performance testing was conducted to evaluate reliability and latency at MQTT quality-of-service (QoS) levels using the one-way delay (OWD) method. The results indicate that the monitoring node’s latency at QoS 0 is relatively stable, averaging 59.51 milliseconds. For the control system, the average latency test was 63.07 milliseconds for QoS 0, 72.21 milliseconds for QoS 1, and 85.64 milliseconds for QoS 2.