BioX Wearables

User Guide

BioX AAL-Band 2.0 wearable sensor and micro-USB charging cable hardware setup
Hardware Ready

BioX AAL-Band 2.0 (fully charged for 6 hours) and the included micro-USB cable.

BioX AAL-Band Bluetooth pairing with Windows laptop for research data collection
PC Specifications

Windows 10/11 operating system with an active Bluetooth 4.0+ receiver.

BioX Desktop Client software interface displaying real-time FMG and IMU sensor data streams
BioX Application

The latest BioX Desktop Client installed and ready to connect.

Table of Contents

BioX FMG Sensor User Guide

Welcome to the official FMG Sensor User Guide for the BioX AAL-Band. This manual covers everything you need to know to set up your wearable hardware, connect to our software application, and start recording high-fidelity Force Myography data.

To download the latest SDK, software, and datasheets, please visit our Downloads Page.

Keep this FMG Sensor User Guide bookmarked for quick reference during your research.

1. Hardware Setup & Placement

Wearing the AAL-Band For optimal signal tracking, place the AAL-Band securely around the upper forearm. Ensure the primary sensor module aligns with the main muscle group you wish to track. The band should be tight enough to track muscle volume changes, but comfortable enough for long-term wear.

Device Controls & LED Indicators The AAL-Band features a minimalist interface with two buttons (Power/Reset and Off) and an LED status indicator:

  • Powering On: Press the On button. The LED will blink red and green twice, indicating the device is active.

  • Resetting Connection: While the device is on, press the On button again. The LED will flash half-green/half-red twice.

  • Powering Off: Press the Off button. The LED will blink red five times before shutting down.

  • No Light: The device is powered off or out of battery.

FMG Sensor User Guide
Figure 2.1: Recommended placement of the AAL-Band on the upper forearm for optimal muscle tracking.
Researcher wearing the BioX AAL-Band 2.0 FMG forearm sensor while typing on a laptop at a desk.
Figure 2.2: The AAL-Band interface, featuring the On/Reset and Off buttons with LED status indicators.
2. Software Connection

Once your hardware is on, it is time to connect to the BioX Application to begin processing your FMG Sensor data.

Connecting to Bluetooth

  1. Turn on your computer’s Bluetooth.

  2. Open your PC settings and click “Add Bluetooth or other device” to pair the AAL-Band.

  3. Open the BioX Application.

  4. Click the Connect button on the start screen. A yellow indicator means it is attempting to connect; a green indicator confirms a successful connection.

    Tip: If the light turns red, reset your AAL-Band or PC Bluetooth and try again.
Bloetooth
Figure 3.1: Pairing the AAL-Band via Windows Bluetooth settings.
AAL-Band User guide
Figure 3.2: The BioX Application start screen.
Figure 3.3a: Yellow indicator showing the application is attempting to connect.
(b) AAL-Band is connected.
Figure 3.3b: Green indicator confirming successful AAL-Band connection.
3. Calibrating the Sensor

This section of the FMG Sensor User Guide ensures your software accurately maps your baseline.

Step-by-Step Calibration

  1. Make a firm fist with the hand wearing the AAL-Band.

  2. Click Calibrate in the application.

  3. Hold the fist for three seconds (the indicator light will be yellow). When the light turns green, calibration is complete and you can relax your hand.

  4. (Optional) If you make a mistake, simply press Reset Calibration and try again.

AAL-Band User Guide
Figure 3.4: Calibration in progress (Hold gesture for 3 seconds while yellow light is active).
Figure 3.5: Calibration complete. Use "Reset calibration" if the gesture was faulty.
4. Recording Data & Training Models

Continuing with our FMG Sensor User Guide, the BioX application allows you to record specific gestures and train a machine-learning model.

Step-by-Step Data Recording

  1. Navigate to the Record tab in the application.

  2. Select your desired recording duration (e.g., 5 seconds per gesture) using the Rec time setting.

  3. Choose the gestures you wish to record. (Note: Your first session must include at least two distinct gestures).

  4. Click Start Recording.

  5. Perform the gestures shown on your screen. A loading bar will indicate how long to hold the pose. Tip: Keep your arm in a consistent position for each recorded gesture to ensure clean data.

  6. Optional: If you make an error, click Reset Data to clear the session and start over.

Tip: To add or remove custom gestures to customize the gesture list, please navigate to the application folder on your PC and open the hand_pic folder. Add images here (named according to the gesture) to include them, or delete images to remove them from the software.
Training Your Gesture Model Once satisfied with your recorded data, choose your data parameters (e.g., FMG, Gravity, Angular Velocity) and click Train Model. The status indicator will turn green once successfully trained.
AAL-Band User Guide
Figure 3.6: The Data Recording menu and gesture selection panel.
AAL-Band User Guide
Figure 3.7: Recording in progress. The loading bar indicates the required duration for the selected gesture.
AAL-Band User Guide
Figure 3.8: Gesture model successfully trained, indicated by the green status light.
5. Testing & Output Data

With your model trained, you can now test its real-time recognition capabilities or export the data.

Testing Your Model

  1. Navigate to the Testing tab (Our software utilizes a Support Vector Machine classifier).

  2. Click Start and perform the gestures you recorded. The application will display an image of the predicted gesture.

  3. If the classifier struggles, return to the Record tab, collect more data for that gesture, and retrain the model.

  4. Note: A live “Output Graph” feature for real-time visualization during recording/training is currently under development.

Data Export & Output Files To save your session data, click File > Save Session. The application outputs .csv files for each gesture (spanning 21 columns) and generates up to 5 distinct dumps:

  • Raw Data

  • Root Mean Square (RMS) of data windows

  • Input for Support Vector Machine (X and Y)

  • Classifier Model (Saved as a .sav file via pickle)

  • Testing Data (File formatting follows: [Date]_[Time]_[DataType]_[Gesture].csv)

Using MATLAB or Python If your primary goal is to collect and plot raw data without our UI, use our open-source scripts on the Downloads Page. Simply adjust the time_to_record (MATLAB) or sampletime (Python) variables to dictate your collection window.

AAL-Band User Guide
Figure 3.9: The Testing menu, utilizing the Support Vector Machine classifier.
AAL-Band User Guide
Figure 3.10: Real-time recognition displaying the predicted "Flexion Vertical" gesture.
Figure 4.1: Placeholder view for the upcoming real-time data output graph feature.
Table 4.1 Description of output datafiles
Table 4.1: Detailed breakdown of the 21 data columns generated in the output .csv files.