This project focused on the design and development of an autonomous drone primarily intended for agricultural applications, with the flexibility to support a wide range of use cases such as autonomous flying , crop monitoring and precision spraying. A key objective of the system was to achieve at the end high payload efficiency, with the drone engineered to carry up  own weight while maintaining stability, control, and reliability during spreading operation.

The system is built around a distributed architecture, where a Raspberry Pi integrated with a CAN (Controller Area Network) HAT serves as the high-level controller. This setup enables robust, real-time communication between subsystems such as motor controllers, sensors, and embedded microcontrollers. The CAN bus ensures fault-tolerant and efficient data exchange, allowing different modules to operate independently while remaining synchronized within the overall system.

To enable autonomous navigation and advanced control, the drone integrates MAVROS, a middleware package that connects ROS (Robot Operating System) with MAV Link-based flight controllers. This allows the Raspberry Pi to communicate directly with the flight controller, enabling high-level commands such as waypoint navigation, velocity control, and mission execution. MAVROS plays a critical role in bridging software intelligence with low-level flight control, making it possible to implement autonomous behaviors.

For environmental perception and obstacle avoidance, the system incorporates a mini LiDAR sensor compatible with ROS1/ROS2 and Raspberry Pi platforms. This LiDAR provides real-time distance measurements, allowing the drone to detect obstacles, maintain safe navigation paths, and potentially perform terrain mapping. By integrating LiDAR data into the control system, the drone can adapt to dynamic environments and improve operational safety.

Additionally, the drone is designed to support mini pump systems for applications such as precision agriculture spraying or fluid delivery. These pumps can be controlled via the embedded system and coordinated through the CAN network, enabling controlled dispensing based on mission parameters or sensor feedback. This adds a functional layer to the drone, extending its capabilities beyond sensing and navigation into active field operations.

On the software side, the Raspberry Pi processes sensor data, runs control algorithms, and coordinates all subsystems. By combining ROS-based frameworks, real-time communication via CAN, and embedded control logic, the system is designed to be modular, scalable, and adaptable to different mission requirements.

My Role

As the Communication Lead, I am responsible for integrating all CAN bus nodes and managing the drone’s communication framework. This includes ensuring reliable, low-latency communication between the Raspberry Pi and all subsystems, coordinating data flow for sensors, actuators, motor controllers, and liquid delivery pumps. Using CAN, I will enable precise control over the mini pumps, allowing the drone to dispense liquids accurately and autonomously while simultaneously managing flight and navigation systems. My work ensures that all hardware and software components operate seamlessly together as a unified, fully autonomous system

Skills 

• Integration and configuration of CAN bus communication systems within distributed embedded architectures

• Application of real-time communication protocols to coordinate sensors, actuators, and control units

• Integration of software and hardware components using Raspberry Pi, ROS/MAVROS, and embedded controllers

• Implementation and coordination of control systems for physical processes, including precise operation of liquid delivery subsystems

• System-level integration and synchronization of multiple subsystems in an autonomous platform

• Debugging, testing, and validation of communication networks in complex, real-world environments

Resources used 

• Raspberry Pi with CAN HAT – for high-level control and CAN bus communication integration

• ROS / MAVROS (MAVLink) – for communication between the onboard computer and flight controller, enabling autonomous operation

• CAN Bus Protocol Documentation – for configuring and managing reliable communication between distributed subsystems

• Mini LiDAR Sensor (ROS-compatible) – for obstacle detection and environmental sensing

• Embedded Controllers / Microcontrollers – for subsystem-level control (motors, pumps, sensors)

• Mini Pump Systems – for liquid delivery and precision spraying functionality