What if a person who cannot speak could still express basic needs using only eye blinks? In this project, we explore the world of Human-Computer Interaction (HCI) and Brain-Computer Interface (BCI) by creating a system that translates eye blinks into spoken words.
Mind to Words (M2W) is a communication system designed to help ALS patients and individuals with limited voluntary movement interact with their environment. Using the Neuro PlayGround (NPG) Lite—a powerful, wireless bio-potential sensor—we capture tiny electrical signals from the eyes (EOG) to control a speech-assistance application.
This is a completely "hands-free" communication tool. By capturing these bio-potential signals, users can hover through a menu of six essential needs—Water, Food, Washroom, Help, Outing, and Television—and select them to be spoken aloud simply by blinking. This low-cost assistive technology provides a vital voice to those who need it most, turning biopotential signals into meaningful human connection
- 1 x NPG Lite
- 1 x Vibz Playmate
- 3 x BioAmp Snap Cables
- 3 x Gel electrodes
- 1 x Nuprep skin preparation gel (Optional )
- 1 x Alcohol swab
- 1 x Laptop
Tip: You can get everything in the Ninja Pack From ROBU.IN, , Upside Down Labs online store , International Customer (Tindie)
Software- NPG Lite Flasher web Application
- M2W Web Application
- Chromium-based browser (Google Chrome, Brave, or Microsoft Edge) for the M2W app.
- Apply a small amount of Nuprep gel (optional)
- Gently rub on electrode placement areas
- Clean with alcohol swabs
This removes oil and dead skin, improving signal quality
Once your skin is prepared, it’s time to connect BioAmp Snap Cables.
- Positive of Channel 1 to A0P
- Negative of Channel 1 to A0N
- Reference cable to REF
Measuring EEG using Gel electrodes:
- Connect the BioAmp Cable to gel electrodes,
- Peel the plastic backing from the electrodes.
- Place the cable connected to A0P between the Fp1 and Fp2 positions
- Place the cable connected to A0N on the bony part behind the earlobe on one ear.
- Place REF on the bony part behind the earlobe of the other ear.
Before connecting the device to the software, we first need to properly power the NPG Lite.
- Connect the battery cable properly to the NPG Lite board to power the device wirelessly. Make sure the connector is securely attached before starting.
- Turn ON the NPG Lite board by switching the ON/OFF button to the ON position. The board is now ready to connect and use.
- Connect NPG Lite to your computer via USB
- Open NPG Lite Web Flasher.
- Click Connect → Select USB JTAG
- Click Get from GitHub → Select “BCI-Blink-BLE”
- Flash firmware
Once done, disconnect USB — the device will work wirelessly.
Step 6: Connecting to M2W App
- Open the M2W Web app in chrome/brave: https://upsidedownlabs.github.io/M2W/
- Turn on Bluetooth (Note:Just turn on Bluetooth and donot manually connect NPG Lite to bluetooth)
- Click Connect to scan for BLE devices.
- In the dialog box, select ESP32C6_EEG (your NPG Lite device).
- Once connected, electrode signals are streamed in real-time.
- Try blinking:
1. Double blink: Hover menu items
2. Triple blink: Speak/select the option
Step 7: How It Works?The system detects blink patterns:
- Single Blink: No action
- Double Blink: Navigate through the menu
- Triple Blink: Select the option and speak it out loud
- Perform strong, deliberate blinks rather than natural eye flutters.
- Sit still, keeping your head steady and jaw relaxed to avoid muscle artifacts.
- Stay at least 5 meters away from heavy appliances (microwaves, fluorescent lights) to reduce electrical noise.
- Ensure firm electrode contact with the skin.
- If the signal is poor, clean the skin, reapply gel, and check placement.
- Verify the BCI-BLINK-BLE firmware is flashed and select ESP32C6_EEG when pairing.
- Keep the NPG Lite within 10 meters of your system.
- If the browser won't scan for Bluetooth, type chrome://flags in your URL bar, enable Web Bluetooth, and restart the browser.
- Current uses include assistive communication, hands-free accessibility, and BCI/HCI education.
- Future upgrades aim to include machine learning classification, custom blink patterns, free-form text-to-speech, and multi-user profiles.

Comments