PAIesque: An Open-Source Cardiovascular Fitness Tracker for Android

An Android app that helps you track and improve your cardiovascular health using your heart rate data. It calculates a personalized activity score and monitors your resting heart rate, giving you clear, actionable insights into your fitness progress - all, with no subscriptions or data sharing.

How It Works:

PAIesque is inspired by public research from the NTNU Cardiac Exercise Research Group (CERG). Their foundational studies on the PAI (Personal Activity Intelligence) metric demonstrated a powerful finding:

Our research shows that those who achieve 100 PAI or more every week over time live on average eight years longer than others.

While PAIesque does NOT use the official PAI® algorithm (that is not known anyway), it applies this same, powerful principle: Your score rises faster when your heart works harder - so both long brisk walks and vigorous workouts move you toward your goal.

Key Differentiators:

• Open & Transparent: Unlike “black box” scores, every calculation is based on the disclosed Karvonen Heart Rate Reserve (HRR) method in combination with a simplified Training Impulse Calculation (TRIMP). You always know how your score is determined.
• Universal & Consistent: Use it with almost any fitness tracker supported by Gadgetbridge (Amazfit, Garmin, Xiaomi, Fitbit, 100+ devices). Your scoring method stays the same even if you change devices.
• Privacy-First & Local: Everything is processed on your phone. Your heart rate data never leaves your device, and no internet connection or account is required.

Note: This is an independent, community-developed application. It is based on publicly available scientific research but is not affiliated with or endorsed by NTNU CERG, and its score is distinct from the proprietary Personal Activity Intelligence (PAI®) metric.

Screenshots

Main Screen	Intensity Zones	Trend	Trend	Devices	Fitness Score Settings	RHR Settings
Summary	Zone Analysis	7-day	30-day	Devices	Heart Rate Reserve	RHR Masurement Time

[format.html](../../dev/format.html)

Installation

1. Prerequisite - Gadgetbridge

a. Install and configure Gadgetbridge

...to read heart rate data from your fitness tracker / wearable.

b. Enable automatic Export

Gadgetbridge	Gadgetbridge	Gadgetbridge	Gadgetbridge	Gadgetbridge	PAIesque
Open Settings	Open Automations	Open Auto export database	Enable + Export Location	Run AutoExport Now	Import into PAIesque

2. For Users: Install the App

Cardiovascular Fitness Tracker is available on F-Droid - the free and open-source app store:

Recommended for most users - automatic updates, verified builds, and easy installation.

As an alternative to F-Droid, you could use apps like Obtainium.

3. For Developers: Build from Source

If you want to contribute, test development versions, or customize the app:

git clone https://codeberg.org/ojrandom/paiesque
cd paiesque
./gradlew assembleDebug

Overview

This application focuses on two key health metrics:

• Cardiovascular Fitness Score: A direct, simplified application of the HRR based Training Impulse (TRIMP) method to model cardiovascular stress linked to long-term mortality risk
• RHR Analysis: Resting heart rate tracking and trends from your device data.

Research-Backed Methodology

This implementation follows research principles from the NTNU Cardiac Exercise Research Group (CERG):

• Primary Research: Based on the Nord-Trøndelag Health Study (HUNT) tracking over 45,000 participants
• Independent Validation: Confirmed in the Aerobics Center Longitudinal Study with 56,000 participants followed for 15 years
• Health Benefits: Maintaining 100+ Points weekly is associated with reduced cardiovascular disease risk

Disclaimer: This is an independent implementation based on published research principles. It is not affiliated with or endorsed by NTNU CERG. For official PAI scores, use authorized applications. PAI® is a registered trademark.

Why Choose This App?

Universal Standard Across Devices

• Device Independence: Get consistent scores regardless of your fitness tracker model.
• Manufacturer-Agnostic: Works with any device supported by Gadgetbridge (Amazfit, Fitbit, Huawei, Xiaomi, and many more).
• Data Consistency: Same calculation method applied to all your historical data.

See this dedicated page for a practical demonstration of the problem.

Open & Transparent

• Open Source: Full visibility into the algorithms and calculations.
• No Black Box: Understand exactly how your scores are calculated.
• Community-Driven: Developed openly with community input and review.

Privacy-First

• Local Processing: All calculations happen on your device—no cloud dependency.
• Your Data Stays Yours: Heart rate data never leaves your phone.
• Offline Capable: Works completely without an internet connection.

Tested Devices & Wearables

This app works with any device supported by Gadgetbridge. The following devices have been specifically tested:

Confirmed Working

• Garmin: Fenix 7 Pro, Vivoactiv 5, HRM-Pro Plus (low resolution data only)
• Amazfit: Bip U, Band 5/7, Active 2 (Round), Helio Strap (low resolution data only)
• Xiaomi: Mi Smart Band 6
• Huawei: Watch Fit 4 Pro
• Colmi: R09, R02
• Pine64 PineTime

🌐 All Gadgetbridge-Compatible Devices

Since the app uses Gadgetbridge for data import, it should work with all 100+ devices supported by Gadgetbridge.

Want to add your device? Test the app with your wearable and:

• open an issue, or
• submit a merge request via git to update this list!

Bonus Privacy

Devices that allow you to disable Bluetooth to prevent tracking:

• Amazfit Active 2 (Round)

Use BLE Radar to check what data your devices broadcast.

Data Flow

Your fitness tracker
    ↓ (read by Gadgetbridge)
Gadgetbridge.db
    ↓ (file copy by app)
gadgetbridge.db (read-only)
    ↓ (data sync by app)  
app.db (read-write)
    ├── Heart rate data
    ├── Fitness score calculations 
    ├── RHR calculations
    └── App settings

Calculation Details

Scientific Foundation

The algorithm translates physiological stress on the cardiovascular system (heart rate over time) into a quantified score. This is based on the established Karvonen (HRR) method, a standard in exercise science for personalizing intensity, and is similar in concept to foundational training load models like TRIMP.

Core Formula

The score is calculated using Heart Rate Reserve (HRR) Intensity:

Heart Rate Reserve

HRR = maxHR - RHR

Intensity

Intensity = (HR - RHR) / (maxHR - RHR)

HR... heart rate
RHR... resting heart rate
maxHR... maximum heart rate
HRR... heart rate reserve

Intensity to Points Mapping

Intensity Band	Points per Minute
< 50% HRR	0.0
50-59% HRR	0.2
60-69% HRR	0.5
70-79% HRR	1.0
80-89% HRR	1.5
≥90% HRR	2.0

Daily and Weekly Calculation

• Daily Points: Sum of all second-by-second points.
• 7-Day Score: Rolling sum of the last 7 days' points.

Core Algorithm of Fitness Score

flowchart TD
    Start[Start with HR Data] --> Pairs[Process HR Point Pairs]
    
    Pairs --> ForEach[For each consecutive pair A→B]
    
    ForEach --> Check{Is time gap valid?<br>0 < seconds ≤ 300}
    
    Check -->|NO<br>Too large or zero| Skip[Skip: Invalid interval]
    
    Check -->|YES<br>Valid range| Math[Math Operations:<br>• HR change per second<br>• Find zone boundaries crossed]
    
    Skip --> NextCheck{More pairs?}
    
    Math --> Split[Split at boundaries<br>Where HR changes zones]
    
    Split --> Distribute[Add time to<br>each zone bucket]
    
    Distribute --> NextCheck
    
    NextCheck -->|YES| ForEach
    
    NextCheck -->|NO| Multiply[Final Calculation:<br>Zone time × points/min]
    
    Multiply --> Result[Output: Daily Points Result<br>with total + breakdown]
    
    style Skip fill:#ffebee,stroke:#c62828
    style Math fill:#e8f5e8,stroke:#2e7d32
    style Check fill:#fff3e0,stroke:#f57c00

How It Works- Example Calculation

flowchart TD
    Profile[User Profile:<br>RHR=60, MaxHR=200] --> Calculate[Calculate Personal Zones]
    
    Calculate --> Zones[Personal Zones:<br>• Zone 0: <130 bpm ─ 0.0 Points/min<br>• Zone 1: 130-144 ─ 0.2 Points/min<br>• Zone 2: 144-158 ─ 0.5 Points/min<br>• Zone 3: 158-172 ─ 1.0 Points/min<br>• Zone 4: 172-186 ─ 1.5 Points/min<br>• Zone 5: >186 ─ 2.0 Points/min<br>]
    
    Zones --> Match[Match HR Data to Zones]
    
    Data[HR Data: 130→160 in 120s] --> Match
    
    Match --> Interpolate[Interpolate: Find when<br>130→144→158→160]
    
    Interpolate --> Distribute[56s Zone2 + 56s Zone3 + 8s Zone4]
    
    Distribute --> Score[Score: 0.787 Points<br>56s@0.2/min + 56s@0.5/min + 8s@1.0/min]

Implementing Core Classes

• PAIesqueCalculator.java
• SimpleRHRCalculator.java

Gadgetbridge Integration

The core calculation class could be a candidate for Gadgetbridge integration. There was a feature request to add similar functionality to Gadgetbridge: Add support to PAI Health.

Why a Consistent Standard Matters

This calculator addresses a gap in wearable ecosystems: consistent health scoring across different devices. Unlike manufacturer-specific solutions, this app provides:

• Cross-Platform Consistency: Same calculation whether you switch devices.
• Historical Comparability: Compare your fitness progress even when changing devices.
• No Vendor Lock-in: Your health data isn't trapped in proprietary platforms.

Similar Products

• Mia Health by NTNU (Note: Uses a newer AQ algorithm)
• Strava's fitness score
• Garmin's Intensity Minutes

Comparison of Fitness Scoring Systems and Methods

1. Comparison of Specific Fitness Scoring Systems

The following table details specific systems, including the community-developed PAIesque app, comparing their formulas, heart rate methods, and practical aspects like transparency and cost.

Feature / System	Heart Rate Zone & Point Calculation	Heart Rate Zone Basis (HRR Clarification)	Primary Purpose	Accessibility, Transparency & Limitations
Personal Activity Intelligence (PAI)	Formula not public. Accumulates points based on heart rate above a personalized threshold over a 7-day rolling period. Points are not linear; higher intensity yields points faster.	Uses a personalized model based on age, sex, resting HR, and max HR. Likely uses Heart Rate Reserve (HRR) as a core component.	Long-term health outcome. Backed by research linking a weekly score of 100+ to reduced mortality risk.	Proprietary/Black Box. Official metric now via Mia Health app (subscription-based). Requires specific compatible devices or app.
PAIesque App	Formula fully disclosed. 1. Intensity = (HR - RHR) / (maxHR - RHR) 2. Points per minute mapped to 6 HRR% bands (e.g., 50-59% HRR = 0.2, ≥90% HRR = 2.0). 3. Rolling 7-day sum.	Based on Heart Rate Reserve (HRR): HRR = maxHR - RHR. Zones are defined as percentages of this personalized reserve (e.g., 50-59% HRR).	Independent, research-based health tracking. Provides a transparent, device-agnostic PAI-like score.	Open Source & Transparent. All calculations are visible. Free, no subscription. Requires technical setup with Gadgetbridge. Device-agnostic.
Fitbit Active Zone Minutes	Points accrue for time spent in 3 zones: • Fat Burn (> 40% HRR): 1x point/min. • Cardio (> 70% HRR): 2x points/min. • Peak (> 85% HRR): 2x points/min.	Uses Heart Rate Reserve (HRR) for personalized zones. Max HR often uses the Tanaka method (208 - 0.7 × age).	Meeting weekly activity guidelines. Goal is ≥150 AZM/week (aligned with WHO).	Proprietary but clear. Zone logic is published. Requires Fitbit device & app (subscription for full insights).
Garmin Intensity Minutes	Points for time in 2 levels: • Moderate (≥ 50% HRR / 64% Max HR): 1x minute. • Vigorous (≥ 70% HRR / 77% Max HR): 2x minutes. Goal is 150+ min/week.	User-configurable basis: 1. % of HRR (Karvonen method). 2. % of Max HR. 3. % of Lactate Threshold (most accurate for performance).	Guideline-based activity tracking & training. Integrates with Garmin's broader training ecosystem.	Proprietary but flexible. User can choose and customize the calculation basis. Requires Garmin device.
Strava Relative Effort	Not zone-based. Calculates a composite score: Relative Effort = (Normalized Graded Heart Rate as % of HRR)² × Duration. Vigorous efforts are exponentially weighted.	Relies on Heart Rate Reserve (HRR) to normalize effort. Requires accurate user profiles for Max HR and Resting HR.	Athletic training load management. Quantifies workout strain to balance training and prevent overtraining.	Proprietary formula. Algorithm is not public. Requires a heart rate monitor and Strava subscription for full analysis.

2. Foundational Exercise Intensity Methods

These are the general, science-based methods used in medical studies and by professionals to prescribe and analyze exercise. The systems above are often built upon these principles.

Method	Core Principle & Formula	Heart Rate Basis	Primary Application	Relation to Scoring Systems
Karvonen Method (HRR)	Calculates Target HR Zone: Target HR = [(Max HR - Resting HR) × %Intensity] + Resting HR. Personalizes intensity based on individual reserve capacity.	Heart Rate Reserve (HRR): HRR = Max HR - Resting HR. This is the standard medical/physiological definition.	Clinical exercise prescription, personalized training zones. The gold standard for personalization.	Foundation for PAIesque, Fitbit, and Garmin's HRR mode. PAI's model is almost certainly based on HRR principles.
Percent of Max Heart Rate	Simpler Target HR: Target HR = Max HR × %Intensity. Uses a fixed percentage of theoretical maximum.	Maximum Heart Rate (HRmax): Often estimated by formulas like 220 - age (less accurate).	General fitness guidance, population-level recommendations.	Garmin offers this as a simpler option. Less personalized than HRR.
Training Impulse (TRIMP)	Quantifies Training Load: Original TRIMP = Duration (min) × ΔHR Ratio × Gender Factor. ΔHR Ratio = (Exercise HR - Resting HR) / (Max HR - Resting HR).	Based on Heart Rate Reserve (HRR) for the ΔHR Ratio component. A foundational model for internal load.	Sports science research, periodization for athletes.	Historical Context: Strava's Relative Effort and PAI's algorithm are sophisticated evolutions of this HRR-based load concept. PAIesque is a direct, simplified application.
Rating of Perceived Exertion (RPE)	Subjective Intensity Scale: The Borg Scales (e.g., 6-20 or 0-10) where you self-report how hard you feel you are working.	Not heart rate based. Correlates with heart rate and lactate. Validates subjective effort.	Clinical settings, complementing HR data. Used for a complete picture when HR monitoring is impractical.	Context: Even precise systems like PAI can be validated against RPE. It's a reminder that internal feeling matters alongside metrics.

3. Key Insights and Summary

• Formula Transparency: PAIesque is unique in providing a fully disclosed formula, acting as an open-source reference implementation. The official PAI and Strava algorithms are proprietary "black boxes," while Fitbit and Garmin publish their zone logic but not their full scoring engines.
• Heart Rate Reserve (HRR) in Practice: While PAIesque, Fitbit, and Strava all use the standard HRR (Max HR - Resting HR), they apply it differently:
  • PAIesque & Fitbit: Use HRR to define fixed percentage zones for linear point accrual.
  • Strava: Uses HRR as a normalization factor in a complex, non-linear, exponential equation.
  • Garmin: Offers the most user choice, including the gold-standard Lactate Threshold for athletes.
• Purpose & Philosophy: Systems are optimized for different goals:
  • Health Outcome (PAI, PAIesque): Use HRR to model cardiovascular stress linked to long-term mortality risk.
  • Guideline Adherence (Fitbit, Garmin): Use HRR/%Max HR to track time spent meeting public health activity recommendations (e.g., 150 min/week).
  • Performance Management (Strava, Garmin Lactate): Use HRR or more advanced metrics to quantify training load and optimize athletic performance, emphasizing high-intensity work.
• Access Models: There is a trade-off between convenience, cost, and control:
  • Commercial/Proprietary (PAI/Mia, Fitbit, Garmin, Strava): Easier setup, integrated hardware/software, but often require specific devices, subscriptions, and lock you into an ecosystem.
  • Open/Community (PAIesque): Free, transparent, and device-agnostic, granting full data ownership. The cost is a more technical setup process (using Gadgetbridge).

Conclusion: Choosing a system depends on whether your primary goal is long-term health (PAI/PAIesque), meeting activity guidelines (Fitbit/Garmin), or athletic performance (Strava/Garmin), and how much you value transparency and data control (PAIesque) versus convenience and integration (commercial platforms). Understanding the foundational HRR and TRIMP concepts reveals how most modern scores are interconnected evolutions of core exercise science principles.

For Developers: Setup & Building

Prerequisites

• Android Studio (latest version recommended)
• JDK 11 or higher
• Android SDK Platform 33 or higher
• Git

Build from Source

Clone the repository

git clone https://codeberg.org/ojrandom/paiesque
cd paiesque

Open in Android Studio

android-studio .

Or build from command line

./gradlew assembleDebug

References & Research

• Heart Rate Reserve Research: NTNU CERG
• Mia Health: by NTNU
• Original Study: Nes et al. (The HUNT Study) - Cardiovascular Health Benefits

License

MIT License

Copyright (c) 2025

This implementation is provided for educational and research purposes based on publicly available scientific principles.

Contributing

Contributions are welcome! Please feel free to submit pull requests or open issues for:

• Bug fixes
• Performance improvements
• Enhanced documentation
• Translations
• Gadgetbridge integration progress

---

Maintainer: OJ Random
License: MIT
Status: Active Development