IKSHANA: A Real-Time Text-to-Braille Convertor

Last summer, I built IKSHANA with two other team members, Vani Bedi and Rudra Gupta, as part of a five-week bootcamp hosted by the Indian Institute of Technology, Delhi. The program gave us access to fabrication facilities like 3D printers and laser cutters, as well as a wide range of tools, components, and lab space that made rapid prototyping possible. We also had excellent mentorship from researchers and engineers who helped guide our design decisions throughout the program. Separately, it was my first time working with a Raspberry Pi – while I had coded in Python before, this was my first experience seeing software directly drive a physical system end-to-end.

Out of that program came IKSHANA, a real-time text-to-braille convertor that can scan text from physical sources like newspapers, postcards, or phone screens and instantly generate braille.

The goal of IKSHANA was to make braille more accessible, portable, and responsive to the real world — not just to pre-entered digital content, but to the physical text that people encounter every day.

Our project consists of two main parts: the scanner and the refreshable braille device (RBD). Through our custom OCR system and camera module (8 MP for clear imaging), the scanner first converts external text into a digital form. Our Pi then translates this digital text into a form understandable by the RBD using our novel translation algorithm. The RBD itself is composed of a matrix of solenoids with custom-printed shafts designed for optimal tactile acuity, allowing braille characters to be displayed in rapid succession.

Key Innovations

• Paper-embossed braille is heavy, expensive, and only useful for a brief period of time, which limits its availability across digital content. IKSHANA can translate content from a wide variety of sources, ranging from newspapers and books to phone screens.
• Traditional braille printers require pre-entered digital text and are costly, time-consuming, and not real-time. IKSHANA does not require any digitally entered text — its OCR system, specially designed for imperfect images, can scan text from physical sources with reasonable accuracy.
• The spatial layout of text, which is important for learning, spelling, algebra, and equations, is not preserved by speech synthesis software. This limitation is addressed by IKSHANA’s refreshable braille system, which physically renders characters using a solenoid-based tactile display.
• Our team created a novel algorithm for translating English text into braille characters.
• IKSHANA is portable and compact, giving it an advantage over traditional braille printers and braille books.

The current version of the project is a proof of concept for a single English character. The final design envisions a keyboard-like matrix of characters with multiple rows and columns that refresh once the user reaches the end of the matrix. Future iterations include support for handwritten text, additional languages, and further miniaturization.

Relevant links: Documentation from Iris National Fair 2023