This is the interactive version of the BrailleBridge research paper, presented in an accessible format. The full project repository is available at github.com/Solez-ai.
High-level system architecture
Every Braille chord the student types travels through a clear, five-stage pipeline before arriving as readable text on the teacher’s screen.1
Student presses a Braille chord
The student presses one or more of the six mechanical dot keys simultaneously. The 180 ms chording window ensures that all fingers in a single intentional chord are collected together, even if they land milliseconds apart.
2
ESP32 translates the chord
The ESP32 reads the pressed keys, encodes them as a compact 6-bit bitmask value, and looks up the result in the active language dictionary — English or Bangla. If the Shift key is held, a second layer of dictionaries is searched for numbers, punctuation, math symbols, uppercase letters, or Bangla vowel signs.
3
Dual output is sent simultaneously
The translated character is transmitted over two independent channels at once: Bluetooth HID (to the student’s paired device) and USB Serial (to the Teacher Software on the teacher’s computer). Neither channel depends on the other.
4
Student hears the text via TTS
On the student’s paired phone, tablet, or computer, the text appears as standard keyboard input. The device’s built-in screen reader speaks it aloud, giving the student immediate audible confirmation of what they just typed — without needing to see a screen.
5
Teacher sees readable output in real time
The Teacher Software receives the character over USB Serial through the browser’s Web Serial API. It displays the translated text immediately, timestamps it, adds it to the student’s exportable record, and updates the live monitoring view — all without the teacher needing to know a single Braille cell.