H2Trac EOX-175.

H2Trac EOX-175 Scaled Model.

Cooperative Agriculture Robots

This project focused on the development of cooperative autonomous agricultural robots using scaled models of the H2Trac EOX-175 tractor platform. The system was designed as a research and testing environment for multi-agent robotics, autonomous navigation, and cooperative control algorithms in agriculture.

The platform combined ROS 2 based robotics software, embedded control systems, and simulation workflows to enable multiple autonomous agents to communicate, coordinate, and operate together in a controlled environment. The architecture included both low-level actuator control and high-level autonomy software running on NVIDIA Jetson hardware.

Development covered mechanical and electrical design, CAD modeling, embedded electronics, sensor integration, and distributed robotics software. The project also included Gazebo-based simulation environments, OptiTrack localization integration, path planning, state-machine control, and cooperative multi-robot behaviors.

The system was designed as a scalable robotics research platform capable of supporting autonomous agricultural experimentation, fleet coordination, and future intelligent farming applications. The project emphasized modular software architecture, simulation-driven development, and reliable real-time robotic operation.

My Tasks

Caption for left video.

Caption for right video.

Autoware.Auto Open-source Software

  • Global Planning
    Implemented route planning using A* based freespace planning methods for autonomous navigation and obstacle avoidance.
  • Trajectory Planning
    Generated smooth and drivable trajectories from global paths while respecting vehicle kinematic constraints and steering limitations.
  • Vehicle Control
    Integrated and evaluated both Pure Pursuit and Model Predictive Control (MPC) path-following controllers for Ackermann steering vehicles.
  • ROS 2 & Simulation Integration
    Connected planning and control modules with Gazebo simulation, localization systems, and custom ROS 2 interfaces for autonomous operation and testing.