1 of 1

Autonomous Drone for Smart Farming

Project Advisor: Preetpal Kang

Introduction

The use of drones in agriculture has revolutionized the farming industry by providing valuable insights into crop health, irrigation management, and yield optimization. According to a recent report by the Market.us, the agricultural drone market will exceed 11.08 billion USD by 2032 from an initial 1.7 billion USD in 2022 with a projected compound annual growth rate of 21.2% between 2023 and 2032.

Project Architecture

Testing

Conclusions

Key References

Acknowledgements

Integration with other systems

We want to integrate our project with other agricultural systems, such as irrigation systems and farm management software, in order to create a comprehensive and automated farming system. This will reduce the need for manual labor and increase total productivity.

Integration with machine learning

Our project has potential to integrate with machine learning for better analysis of the data collected from the sensors and provide real-time recommendation for agricultural activities.

Adding more sensors

We are working towards including other sensors such as multispectral and hyperspectral sensors which can provide comprehensive data about plant health, nutrient content, and disease detection.

We would like to express sincere gratitude to the project advisor, Preet, for his guidance, support and expertise throughout this project. Preet’s assistance and feedback have been a significant influence in the success of this project

Our project aims to take this a step further by utilizing a drone autonomously to create a smart farming system that collects soil moisture and temperature data from “sensor pods” placed at various locations throughout a field. The drone follows a predefined path of GPS coordinates to stop at each sensor pod. This data is then transmitted to a ground station and subsequently to a mobile application for easy access and visualization. Our system provides farmers with a comprehensive understanding of their fields, enabling them to make informed decisions about crop management and resource allocation.

2. Sensor Pod

A sensor pod is equipped with a moisture and a temperature sensor, an ESP32 microcontroller, a 3.7V LiPo battery, a 6V solar panel, and a solar power manager board. The data collected by the sensors is transmitted by the ESP32 to a separate ESP32 located on the drone through the ESP-NOW protocol, allowing for data transmission at ranges up to 500 ft. Debug LEDs were wired on both ESP32 boards to allow for visual indication of system status. For the simplicity, only one sensor pod was assembled for testing.

The sensor pod can perform its functions indefinitely during the daytime as long as the solar panel is not obstructed. This design limits the need for physical interaction after initial installation.

To ensure reliability and accuracy, extensive testing was performed on our system. The testing phase included the following:

1. Flight Testing

Before involving the sensor pod, the drone was run on typical flight paths with varying stopping points along its route. This gave experience allowing for fine-tuning the movement necessary for getting in horizontal and vertical range of the sensor pod.

2. Sensor Pod Testing

Standalone communication testing was performed between the drone ESP32 and the sensor pod before attempting it in flight. These tests were used to determine the minimum required range data could be transferred over with no obstructions. Separate tests helped determine the range required for a stable connection between the drone ESP32 and the ground control station over Bluetooth. After establishing connection between all endpoints, tests were run to ensure that measurements originating at the sensor pod were accurately being transferred to the ground control station.

3. Integration Testing

With both halves of the project functioning, integration tests were conducted to observe the entirety of the system. The ground control station was placed at one end of a large open park with the sensor placed at the opposite end. The drone ESP32 was mounted to the drone and the drone was supplied with GPS coordinates at various spots around the perimeter of the park, including the sensor pod.

Project Architecture

3. Mobile Application

When the drone returns to its original takeoff location, it deposits the sensor data it collected on its flight path using a Bluetooth connection between the drone ESP32 and a ground control station. This ground control station uploads this data to a Google Firestore database, which is read by our mobile app. As new data is added to the database, the app reflects this with two separate charts for temperature and humidity. An option to view data averages from multiple flights is also possible.

Computer Engineering Department

Choksi, Dhruv

(MS Computer Engineering)

Doctolero, Jonathan

(MS Computer Engineering)

Zunino, Anthony

(MS Computer Engineering)

Hatzikokolakis, Michael

(MS Computer Engineering)

Future Work

Autonomous Drone

Our project uses an autonomous drone equipped with a Pixhawk flight controller and PX4 autopilot firmware. The flight controller is responsible for drone movement and stability. PX4 autopilot is an open-source software stack with wide industry adoption. It is capable of handling a large range of hardware and can be customized to suit different needs.

We believe that the use of this project in agriculture can greatly improve the efficiency and effectiveness of farming practices, paving the way for a more sustainable and prosperous future. We hope that our project will inspire more research and development in this area and lead to a wider adoption of these technologies in agriculture.

[1] Ammad Uddin, M., Ayaz, M., Aggoune, E. M., Mansour, A., & Le Jeune, D. (2019). Affordable broad agile farming system for rural and Remote Area. IEEE Access, 7, 127098-127116. doi:10.1109/access.2019.2937881

[2] Bono Rossello, N., Carpio, R. F., Gasparri, A., & Garone, E. (2022). Information-driven path planning for UAV with limited autonomy in large-scale field monitoring. IEEE Transactions on Automation Science and Engineering, 19(3), 2450-2460. doi:10.1109/tase.2021.3085365

[3] Caruso, A., Chessa, S., Escolar, S., Barba, J., & Lopez, J. C. (2021). Collection of data with drones in precision agriculture: Analytical model and Lora Case Study. IEEE Internet of Things Journal, 8(22), 16692-16704. doi:10.1109/jiot.2021.3075561

[4] Gupta, M., Abdelsalam, M., Khorsandroo, S., & Mittal, S. (2020). Security and privacy in Smart farming: Challenges and opportunities. IEEE Access, 8, 34564-34584. doi:10.1109/access.2020.2975142

The firmware runs on the microcontroller located on the flight controller board, which receives inputs from various sensors and adjusts the drone’s motor accordingly.

Results

The drone was sent on a complete path and allowed to return to the ground control station. Upon landing, it was observed that the latest measurements from the sensor pod across the park were successfully reflected in the mobile application.