Domain / Technologies

Technologies Used

The stack below reflects the full RP implementation from edge AI and robotics to resilient communication control.

AI and Software Stack

YOLOv11 Python Docker Edge Inference

Embedded and Robotic Stack

Raspberry Pi 5 ESP32 4-DOF Arm 3D-Printed Gripper Sprayer Nozzle GSM Module

Communication and Control Stack

LoRa Mesh ONOS SDN Open vSwitch Adaptive QoS IPsec VPN Failover Logic

Operational Capabilities Enabled

  • Real-time ripeness and disease diagnosis
  • Targeted robotic harvesting and spraying
  • Low-latency edge decision execution
  • Reliable operation under unstable links

Measured Runtime Indicators

  • Ripeness model mAP@0.5: 0.937
  • Disease model mAP@0.5: 0.636
  • Average edge inference latency: ~42ms
  • LoRa mesh PDR: above 98%
  • SDN failover latency: ~165ms

Back to Domain Hub