OrchardEyes

Home | Our drone flying in small orchard
Custom Autonomus Drone with Obstacle avoidance and Route planning equipped with RGB & NOIR cameras
Yield Estimatation, Pest Infestation data Reports
User Dashboard, Infected Area map, suggestions
Quick Actions, Farm insights, Gen AI Rag Chatbot with multilingual and voice support
ML models Results for yield count, disease classification, tree detection & plant part detection
Farms Virtual Tour & Marketing page

Inspiration

The inspiration behind OrchardEyes stems from the inefficiencies in traditional orchard management practices. Manual methods are time-consuming, error-prone, and often lead to suboptimal yields, increased wastage, and higher operational costs. The lack of precise, real-time monitoring systems exacerbates challenges in pest management and irrigation. Recognizing these issues, we aimed to create a solution that leverages modern technology to empower farmers with data-driven insights, enabling them to make informed decisions and improve their yield and sustainability.

What it does

OrchardEyes is an autonomous drone-based system designed to revolutionize orchard management. It utilizes computer vision, AI, and blockchain technology to provide real-time insights into crop health, resource usage, and traceability. The system enables farmers to:

Monitor Crop Health: Detect diseases, pests, and growth issues through advanced image analysis.
Optimize Resource Usage: Recommend efficient irrigation and fertilization schedules based on real-time data.
Enhance Traceability: Use blockchain technology to provide transparent tracking of produce from tree to market.
Improve Efficiency: Reduce labor costs by automating drone surveys and data collection.
Ensure Access for All Farmers: Offer an affordable, scalable solution through a pay-per-use drone service rental system.

How we built it

OrchardEyes is built using a combination of cutting-edge technologies and innovative approaches:

Multi-Camera Image Capture: The drone is equipped with digital, multispectral (NOIR), and thermal cameras to capture comprehensive tree images.
Health Metric Analysis: Captured images are processed to calculate key metrics such as chlorophyll content, disease and pest infestation levels, water stress, and yield estimation.
Selective Data Processing: A YOLO model identifies leaves, flowers, and fruits within tree images, sending only relevant data to the backend for processing, reducing internet usage for farmers.
Efficient Tree Detection: A tree detection model ensures that only one tree is captured per image to avoid redundant data collection.
Cost-Effective NDVI Calculation: Instead of using an expensive full multispectral camera, a combination of NOIR and digital cameras calculates NDVI, providing high accuracy at a significantly lower cost.
Blockchain Integration: Blockchain technology provides farm ratings based on drone-collected data, ensuring traceability and transparency.

Challenges we ran into

Data Processing Efficiency: Ensuring that only relevant data is processed and transmitted to reduce internet usage for farmers.
Cost Management: Developing a cost-effective solution that remains accessible to small and medium-scale farmers.
Precision in Navigation: Achieving precise drone navigation and alignment with trees while avoiding obstacles in real-time.
Integration of Multiple Technologies: Seamlessly integrating computer vision, AI, blockchain, and autonomous navigation into a cohesive system.

Accomplishments that we're proud of

Innovative Solution: Developing a comprehensive system that addresses multiple challenges in orchard management.
Cost-Effective Technology: Creating a high-accuracy NDVI calculation method using affordable camera technology.
User Accessibility: Ensuring the system is accessible to farmers through multiple platforms, including web, mobile, and WhatsApp.
Real-Time Insights: Providing farmers with real-time, actionable insights to improve their yield and sustainability.

What we learned

Importance of Data-Driven Decisions: The critical role of real-time data in making informed decisions for orchard management.
Technology Integration: The challenges and rewards of integrating multiple advanced technologies into a single system.
User-Centric Design: The importance of designing solutions that are accessible and user-friendly for farmers with varying levels of technical expertise.

What's next for OrchardEyes

Scalability: Expanding the system to cover larger orchards and different types of crops.
Enhanced AI Models: Continuously improving the AI models for better accuracy and faster processing.
Farmer Training Programs: Developing training programs to help farmers fully utilize the system and its features.
Partnerships: Collaborating with agricultural organizations and governments to promote the adoption of OrchardEyes.
Research and Development: Investing in R&D to explore new features and technologies that can further enhance orchard management.

🛠️ Tech Stack

Category	Technologies
Frontend	React + Vite, TailwindCSS, React Native (Mobile App)
Backend	Flask (Python), Express.js (Node.js), PostgreSQL, Pinecone/Chroma DB (Vector Storage)
AI/ML	PyTorch, OpenCV, YOLO family models, Weights & Biases (Model training)
Blockchain	Ethereum Sepolia Network with hardhat