Loading System...
Graduation Project (A+)
Farmer Eye
Smart Agricultural Robot for Disease Detection & Monitoring
The Problem
- Manual plant health monitoring is time-consuming and prone to human error.
- Inefficient use of resources (water, fertilizers) due to lack of real-time data.
- Critical factors like soil temperature and wind often remain undetected.
- Delayed identification of diseases leads to significant crop yield losses.
- Absence of data-driven tools limits farmers' ability to make timely decisions.
The Solution
Farmer Eye is an AI-driven autonomous vehicle ecosystem. It navigates farm fields collecting real-time data through high-resolution cameras and sensors. It analyzes plant health metrics using Machine Learning to detect diseases, assess pest infestations, and provide actionable insights via a user-friendly mobile app, ensuring sustainable and precise agriculture.
System Architecture
A cohesive integration of Robotics, AI, IoT, and Mobile Software.
1. Smart Vehicle (Robotics)
- Chassis & MobilityCustom-built RC car chassis with high-torque motors driven by an Arduino microcontroller.
- Sensors & GPSEquipped with DHT sensors for detecting temperature/humidity and a GPS module for precise geolocation of every scanned plant.
- SustainabilityIntegrated Solar Panels provide a reliable, clean energy source to charge batteries and extend operation.
2. AI & Cognitive Core
- Inference OptimizationConverted the model to TensorFlow Lite to enable low-latency, efficient real-time inference on the Raspberry Pi 4 edge device.
- Vision System8MP Camera Module captures high-definition images of crops for analysis.
- Deep Learning ModelLeveraged Transfer Learning by fine-tuning pre-trained CNN architectures on the PlantVillage Dataset.
- ✓ Accuracy: ~94%
- ✓ Loss Function: categorical_crossentropy (38 Classes)
3. Mobile Command Center
- Flutter AppA cross-platform app providing farmers with diagnostic reports, disease location maps, and treatment recommendations.
- Real-time IoTSeamlessly connected to the vehicle via APIs and Firebase/Firestore for live monitoring and history tracking.
4. Data Engineering
- Preprocessing Deep DiveApplied rigorous Normalization (rescaling pixel values to 0-1) and resizing to fixed 224x224 dimensions for consistency.
- Advanced AugmentationEmployed strategies like Rotation, Shear, Zoom, & Shift to synthetically expand the dataset and improve model generalization.
- DatasetUtilized 54,000+ labeled images covering healthy and diseased states of crops like Tomato, Potato, and Apple.
My Role & Responsibilities
I spearheaded the development of the system's cognitive core, acting as the AI Architect. My primary focus was constructing the "brain" of the rover:
- Model Architecture DesignDesigned and implemented specific Deep Learning architectures (CNNs) using Keras and TensorFlow to maximize detection accuracy on edge devices.
- Edge OptimizationOptimized heavy neural networks to run efficiently on the Raspberry Pi 4, ensuring low-latency inference without internet dependency.
- Data Engineering PipelineEngineered robust data pipelines for preprocessing real-time camera feeds and handling the massive PlantVillage dataset for training.
Technology Stack
PythonTensorFlowKerasOpenCVRaspberry Pi 4ArduinoFlutterFirebaseIoTGoogle Maps APISolar EnergyPythonTensorFlowKerasOpenCVRaspberry Pi 4ArduinoFlutterFirebaseIoTGoogle Maps APISolar Energy
Impact & Results
- High AccuracyAchieved 94% classification accuracy on the PlantVillage dataset using fine-tuned CNNs.
- Real-Time InferenceOptimized performance on Raspberry Pi with inference times of ~200ms per image.
- Comprehensive CoverageSupports detection of 38 distinct plant disease classes across multiple crop species.
- Efficiency BoostReduced manual crop inspection time by approximately 70% through automation.
Challenges & Learnings
Edge Deployment Constraints
Running complex CNNs on a Raspberry Pi 4 was computationally expensive. Solution: We utilized TensorFlow Lite quantization to reduce model size by 75% without significant accuracy loss, enabling smooth performance on limited hardware.
Real-time Inference Lag
Balancing frame rate with detection accuracy was critical. Solution: Optimized the inference pipeline to process frames at ~5 FPS, sufficient for a slow-moving agricultural rover, ensuring no diseased plants were missed.
Handling Class Imbalance
The dataset had unequal samples for some disease classes. Solution: Applied advanced data augmentation (Shear, Zoom) and class weighting during training to ensure the model generalized well across all 38 categories.
IoT + AI Integration
Synchronizing the rover's movement with cloud data streaming was complex. Solution: Architected an asynchronous event loop using Python's asyncio to handle motor control and Firebase updates concurrently without blocking the AI inference thread.
Achievements & Recognition
ASRT Funding
Secured funding from the Academy of Scientific Research and Technology to support system development.
Innovation Exhibition
Participated in the 3rd International Youth AI Forum.
ITAC Grant
Awarded funding from ITAC to advance technical research and optimize the intelligent model.
Project Credits
Farmer Eye Graduation Team
Development Team
Supervised By
Dr. Saeed MohsenSupervisor
Eng. Ahmed FaroukSupervisor