Smart Farm Infrastructure with Metaverse Using Computer Vision, oneM2M Communication and Jetson Orin Nano
Specifications
- Code
- RA2616
- Department
- DKR
- Technology
- Jetson Orin Nano · NVIDIA Jetson · oneM2M · Metaverse · Computer vision
- Theme
- Smart Farming · AI & Computer Vision
- Student
- Harith Saufi bin Zairul Adha
- Supervisor
- Noor Nadiah binti Mohd Azali
Objectives
- To design and integrate a multi-sensor farm monitoring system using ESP32 boards with soil moisture, humidity, temperature, light intensity, and air quality sensors to collect real-time environmental data.
- To design and develop an integrated IoT communication system that enables realtime
- interaction between physical farm devices (water pumps, lighting, fans) and a Roblox-based virtual farm using oneM2M and MQTT protocols.
- To implement an intelligent smart farming platform using NVIDIA Jetson Nano and Mobius CSE architecture to perform plant health analysis through computer vision and synchronize real-world farm data with its metaverse representation.
Scope of project
- Smart Farm Hardware — A compact physical farm prototype using ESP32 microcontrollers for real-time environmental sensing and Jetson Nano for computer vision-based plant health monitoring and growth analytics.
- Metaverse Interface — An immersive Roblox environment that acts as a remote control panel with live sensor visualization, bidirectional actuator control, and a virtual farm synchronized with the physical state.
- Integrated Architecture — An end-to-end IoT ecosystem using oneM2M/MQTT via
- Mobius CSE for secure cloud communication, MySQL on Jetson Nano for data logging, and real-time sync between sensors, cloud middleware, and the virtual interface.
Project description
This project develops an intelligent digital twin smart farming platform bridging physical agriculture with an interactive metaverse environment. Utilizing ESP32 microcontrollers, the system continuously monitors vital environmental parameters (soil moisture, temperature, humidity, light, and air quality) and regulates automated actuators including irrigation pumps, grow lights, and ventilation fans. At the edge, an NVIDIA Jetson platform performs computer vision-based plant health and growth analytics. Data exchange and device interoperability are standardized through the oneM2M architecture via Mobius CSE and MQTT protocols, backed by local logging. The system synchronizes physical farm telemetry in real-time with an immersive Roblox-based virtual twin, enabling bidirectional control, remote visualization, and intuitive precision agriculture management.