What the app does
What it does
Real‑time hydration guidance powered by your Garmin device, personalised to you. Fluid‑Loss Companion estimates your sweat/fluid loss during workouts and builds a personalised model from your past activities to improve accuracy over time.
Who it's for
- Cyclists and triathletes training on the road, TT, gravel, or indoors
- Riders who want actionable hydration guidance during races and long rides
- Optimised for Garmin bike computers via Connect IQ
Key benefits
Accurate
Improve predictions with your own ride history and measured fluid loss.
Real‑time
Live estimates sent to your Garmin bike computer every ~2 minutes.
Simple and focused
No complex setup—pair your device, train your model, and ride.
How it works
Data sources
- Live ride metrics from your Garmin bike computer during workouts.
- Past rides imported from Garmin Connect.
Models
- Generic model: Default predictions without personalisation.
- Personalised model: Trained on your own rides. The app can auto‑choose the best model or you can pick.
Under the hood
The app securely calls a lightweight cloud service to get your current estimate and to train your personal model.
Prediction vs Training
Prediction (during workout)
- Pick your Garmin device in the app and start tracking.
- The app receives metrics and calls the cloud predict endpoint every ~2 minutes.
- The fluid‑loss estimate is shown on the phone and sent to your Garmin device.
- Model choice: If you've trained a personalised model and your preference is Auto or Always Personalised, the app uses it; otherwise it uses the generic model.
Training (between workouts)
- Import past activities by sharing or pasting a Garmin Connect URL.
- Enter your measured fluid loss (mL) for each activity and save.
- Once you have enough activities (≈10+), tap Train Cloud Model.
- Your personalised model is stored server‑side and used for future predictions.
Train your model
Walk through the full training workflow — import rides, add your measured fluid loss, and build a model that learns how you sweat.
- Import past rides from Garmin Connect
- Enter your measured fluid loss for each ride
- Tap Train Cloud Model when you have 10+ rides
- Your personalised model is used for future predictions
Interaction with your Garmin device
Device selection
Connect your Garmin bike computer inside the app. Android may request Bluetooth and Location permissions for BLE scanning.
Live metrics
The app receives real‑time ride metrics via the Garmin Connect IQ Companion channel while you work out.
Feedback to head unit
The latest fluid‑loss prediction is sent to your Garmin head unit periodically, so guidance stays visible on your bars.
Download Now
Get the app completely free for a limited time only.
Get the Companion App
Get the Garmin Data Field
Data and privacy
- Local storage: Recent activities are saved on your device.
- Cloud: Predictions and training happen via a remote API. Your identity is tied to your paired device ID—no personal profile is required in‑app.
Frequently Asked Questions
FluidLoss works with Garmin Edge bike computers including Edge 530, 540, 830, 840, 1030, 1040, and Explore series.
We recommend at least 10 rides with measured fluid loss across different conditions (temperature, intensity, duration) for an accurate personalised model.
Yes. Your activity data is stored locally on your device. Cloud predictions use your device ID—no personal profile is required. We never sell or share your data.
The app is currently available free for a limited time. This includes personalised ML model training, advanced analytics, and all premium features. A Pro subscription will be available in the future.
The app is free for a limited time, including all Pro features. Take advantage of this offer while it lasts.
No extra sensors required. FluidLoss uses existing data from your Garmin device (heart rate, power, temperature, duration) to estimate fluid loss.
Personalised models trained on your own data can approach the accuracy of weighing yourself before/after rides, especially with 15+ training activities across varied conditions.