Talon is a modern, privacy-focused Android application for tracking body composition and fitness progress. Designed to work with Bluetooth smart scales, it provides detailed insights into your health metrics without relying on cloud services.
- Scientifically Calibrated: Body Fat algorithm derived from 59,430 clinical profiles (CDC NHANES dataset).
- DEXA-Grade Accuracy: Our formula correlates raw Bio-electrical Impedance (BIA) with medical Dual-Energy X-ray Absorptiometry (DEXA) scans, replacing generic commercial approximations.
- Direct BLE Connection: Reverse-engineered support for "OKOK" protocol scales (and compatible generic chipsets).
- Ad-Free & Account-Free: No cloud servers, no sign-ups, no ads. Your biometric data stays on your device.
- Google Fit Integration: Seamless sync using the modern Android Health Connect API.
Track more than just weight. Talon captures and analyzes:
- Weight & BMI (with color-coded categories)
- Body Fat %
- Muscle Mass
- Water %
- Bone Mass
- Metabolic Age
- BMR (Basal Metabolic Rate)
- Impedance
Most generic smart scales use obscure, proprietary formulas to estimate Body Fat %. Talon takes a transparent, data-driven approach.
We trained a Linear Regression Machine Learning model on the NHANES 1999-2004 dataset (National Health and Nutrition Examination Survey) provided by the US CDC. This specific period is considered the "Golden Age" of body composition data, as it contains both:
- Input: Raw BIA Resistance (Ohms) - what your scale measures.
- Ground Truth: Whole-body DEXA Scans - the medical gold standard for body fat.
The Result: A custom formula optimized on 59,430 real human profiles to translate your scale's electrical resistance into meaningful health metrics with minimized error.
Formula logic:
BodyFat% = f(Height²/Impedance, Weight, Age, Gender)
- Interactive Charts for all metrics (powered by Vico)
- Filter data by Week, Month, 3 Months, Year, or All Time
- View key statistics: Average, Minimum, Maximum, and Change over time
- Set personalized goals for Weight, Body Fat, or Muscle Mass
- Track progress with visual indicators
- Set deadlines and receive achievement alerts
- Offline First: All data is stored locally on your device using Room Database.
- Export: Export your data to CSV for analysis in spreadsheets.
- Backup: Create full JSON backups of your history and settings.
- Edit/Delete: Full control to modify or remove incorrect measurements.
- Dark/Light Theme support
- Reminders: Schedule daily weigh-in notifications
- Onboarding: Guided setup for new users
| Measure Tab | Scanning UI |
|---|---|
 |
 |
| Empty State | Navigation |
 |
 |
| Settings | Help |
 |
 |
- Android: Kotlin, Jetpack Compose (Material 3), MVVM, Room Database.
- Bluetooth: Android BLE Scanner & GATT.
- Data Science: Python (pandas, scikit-learn) for algorithm calibration
- Language: Kotlin
- UI: Jetpack Compose (Material 3)
- Architecture: MVVM (Model-View-ViewModel)
- Local Storage: Room Database (SQLite)
- Asynchronous: Coroutines & Flow
- Charts: Vico
- Background Tasks: WorkManager
- Navigation: Compose Navigation Suite
The project follows a layered architecture based on Android best practices, separating concerns into UI, Data, and Domain layers:
com.aquilesorei.talon
├── ui/ # Presentation Layer
│ ├── screens/ # Composable screens (Home, History, etc.)
│ ├── components/ # Reusable UI components
│ └── theme/ # Theme definitions
├── domain/ # Domain Layer
│ ├── models/ # Business objects (Measurement, UserProfile)
│ └── usecases/ # Business logic
├── data/ # Data Layer
│ ├── local/ # Local storage (Room DB, DAOs, Entities)
│ └── repository/ # Single source of truth for data
├── viewmodels/ # State holders
├── workers/ # Background tasks (WorkManager)
└── utils/ # Utility classes
- Android Studio Ladybug or newer
- JDK 17+
- Android Device with Bluetooth LE support (min SDK 31 recommended)
- Clone the repository:
git clone https://github.com/Aquilesorei/talon.git- Open the project in Android Studio.
- Sync Gradle files.
- Build and run on your device or emulator.
Note: Bluetooth scanning requires a physical device. Emulators cannot scan for Bluetooth Low Energy devices.
- Connect Scale: Go to the Measure tab and tap the Bluetooth icon to scan for your supported smart scale.
- Weigh In: Stand on the scale. The app will automatically capture your metrics.
- Track: Switch to the Charts tab to see your trends.
- Set Goals: Use the Goals tab to define your targets.
- Settings: Customize your theme and notification preferences in Settings.
Contributions are welcome! Please feel free to submit a Pull Request.
This project is licensed under the MIT License - see the LICENSE file for details.