AI for Smart Farming: Detecting and Identifying Cattle from Drone Images
Project Overview
This project built the first piece of a smart farm in Nottingham. We focused on identifying and tracking individual cattle, creating a setup that can scale into a complete monitoring system in the future. To do this, we built a complete machine learning pipeline that pre-processes 4K drone images, detects the cows, and re-identifies them. The entire pipeline was designed to run efficiently on high-performance computing (HPC) systems.
What were the key results of your research project?
- Developed an end-to-end machine learning pipeline that processes raw 4K drone imagery to detect and identify individual cattle.
- Achieved an 82% success rate in correctly re-identifying the detected cows.
- Engineered the architecture from the ground up to easily scale into a complete, farm-wide monitoring system.
GitHub page: https://github.com/Eyad-Ahmed/n8cir-smart-farming/
How do you feel you have benefitted from completing this internship and has it made you consider future career paths?
Before this internship, my background was rooted in software engineering and full-stack web development, but I had zero experience with machine learning. Over the past eight weeks, with the hands-on guidance of my supervisor, Mingqi Gao, I was able to dive deeply into the field. I can now confidently say I have practical, up-to-date experience building a large-scale ML pipeline from start to finish. Ultimately, this project shifted my trajectory: while I previously only considered a career in web development, I have discovered a genuine passion for machine learning that I want to pursue.
Download presentation slides
Internships 2026 - Eyad AAA Mohamed