Zeytin is an AI-powered mobile assistant designed to empower olive farmers and enthusiasts. Developed with a focus on the rich olive heritage of the Aegean region, Zeytin uses on-device Machine Learning to provide real-time insights into olive health and harvest readiness.
Designed in Denizli, Zeytin is built to be 100% offline, ensuring farmers can use it even in the most remote groves.
- 🔍 AI-Powered Scanning:
- Live Scan: Real-time classification of olives using the device camera.
- Photo Scan: Analyze existing photos for deep health inspection.
- 📉 Harvest Roadmap:
- Ripeness Tracking: Identifies stages from Green to Veraison to Black.
- Quality Insights: Guidance on whether olives are best for premium liquid gold (oil) or traditional brining.
- 🪲 Pest & Disease Intelligence:
- Detects common threats like Olive Fruit Fly or Peacock Spot.
- Provides detailed symptoms and mitigation strategies.
- 🇹🇷 Local Expertise:
- Built-in database for famous Turkish varieties: Gemlik, Ayvalık (Edremit), Memecik, and Domat.
- 🛠 Polished UX:
- Smooth SwiftUI interface with responsive design for iPhone and iPad.
- Tactile haptic feedback for a premium feel.
- Language: Swift 6.0
- Framework: SwiftUI
- Machine Learning: CoreML (Custom
OliveClassifiermodel) - Vision: AVFoundation for real-time camera processing
- Platform: iOS 16.0+
Since this is a Swift Playground App (.swiftpm), you can open it in multiple ways:
- Download or clone this repository.
- Open
Package.swiftor the project folder directly in Swift Playgrounds. - Tap "Run" to start exploring.
- Open the project folder in Xcode.
- Xcode will automatically resolve the package dependencies.
- Select your iPhone/iPad or a Simulator (iOS 16+) and press Cmd + R.
- Launch: Start at the Welcome screen and tap "Start Analysis".
- Scan: Use the Live Scan to point your camera at an olive branch.
- Analyze: The AI will instantly tell you the ripeness level or identify if a pest is present.
- Roadmap: Check the Harvest Roadmap to see your overall harvest readiness percentage based on your scan history.
- Report: If pests are detected, a red alert button appears, leading you to a detailed Pest Report.
Zeytin is designed with privacy at its core. All processing happens on-device. No images, location data, or scan results are ever sent to a server.
Developed by Ahmad Khaled Samim for the WWDC26 Swift Student Challenge.
Designed and Developed in Denizli, Turkey.
