Skip to content

Autonomous UAV Crop Monitoring Platform for NXP HoverGames 3

License

Notifications You must be signed in to change notification settings

rotary-auav-ui/cigritous

Repository files navigation

Welcome to Cigritous Repository!

'cigritous logo'

Cigritous is a NXP Hovergames 3 project by Vishwakarma Aerial Dexterity Research Group Universitas Indonesia for reducing yield losses from pests and weather by utilizing automated drone responder and remote monitoring system.

For detailed concept, please read: Crop Monitoring with Automated UAV Spray Response

Authors

  • Thariq Hadyan (Electrical Engineering 23)
  • Benedicto Matthew W. (Computer Engineering 24)
  • Raihan Syahran (Electrical Engineering 23)
  • Abdul Fikih K. (Computer Engineering 25)
  • M. Rizky Millennianno (Physics 23)
  • Dylan Vieri (Computer Engineering 25)
  • Ricky Iskandar Z. (Physics 24)
  • Lauren Christy T. (Computer Engineering 25)
  • Juan Jonathan (Computer Engineering 25)
  • Muhammad Fikri (Mechanical Engineering 24)
  • Andre Christoga P. (Physics 24)
  • Ahmad Rifqi (Computer Engineering 25)
  • Adrian Leo P. (Electrical Engineering 25)
  • M. Daffa Aryasetya (Electrical Engineering 25)
  • M. Fikri R. Abyadhi (Electrical Engineering 25)
  • Rizky Awanta Jordhie (Computer Engineering 25)

Branch Information

Onboard Program Branch

main

Branch for drone computer programs. Contains drone visual-inertial, pest detector, and precision landing program

Hardware:

  • NXP Vehicle Drone Kit (RDDRONE-FMUK66)
  • NXP i.MX 8M Plus based 8M NavQ Plus Computer
  • Google Coral 5MP Camera

Software:

  • Linux Ubuntu 20.04 LTS Focal Fossa / 22.04 LTS Jammy Jellyfish
  • ROS2 Foxy Fitzroy / Humble Hawksbill
  • PX4-Autopilot v1.13.3

Installation and Usage

Click Here

Additional Files

  • AprilTag

Directory: docs/apriltags

We use AprilTag with 36h11 family

  • Crow Image (for testing purposes)

Directory: docs/crow_image

Detection Performance using NXP NavQ+ Single Board COmputer

'crow detected'

Model: Quantized INT8 YOLOv5 Average detection time: 2800ms or 2.8s 'inference performance'

Sensor Network / Ground Module Branch

ground-module

Branch for sensor network or ground module programs. Contains sensor nodes source codes and libraries.

The sensor network has 2 main part: central and node. The nodes are connected together by mesh and connected to sensors to sample ground data. Sampled data will be sent to central, which acts as 'back end' and communicating to onboard computer and FCU via TCP & MAVLink.

Hardware:

  • Bosch Sensortec Environmental Gas Sensor BME688
  • ESP32 DevKit
  • YL-38/69 Moisture Sensor
  • DHT22 Humidity and Temperature Sensor

Software:

  • Arduino IDE 2.x or 1.x
  • Libraries (contained in libraries folder):
    • painlessMesh
    • TaskScheduler
    • DHTesp
    • YL3869
    • MQUnifiedsensors
    • arduino-mqtt
    • Bosch BME68x library

Installation and Usage

Read Me!

Web Server Branch

web_server

Branch for webserver front end. Acts as Web-based Ground Control Station which display agriculture and drone data from sensor network. Tested in Heroku.

Installation and Usage

Read Me!

About

Autonomous UAV Crop Monitoring Platform for NXP HoverGames 3

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published