Towards a "free" braille display

A proposal for a fun and useful microcontroller project

Mario Lang

Outline

  • What is braille and how do braille displays work
  • A closer look at a few details
  • Your participation is needed
  • Discussion

What is Braille?

  • Tactile (non-linear) writing system (uses only dots) for the blind
  • Invented by Frenchman Louis Braille around 1821
  • Originally with 6 dots (2 by 3): ⠿
    • 2^6 = 64 possible combinations
  • For computers extended to 8 dots (2 by 4): ⣿
    • 2^8 = 256 possible combinations

Braille is a non-linear writing system for the blind (and visually impaired) invented by the Frenchman Louis Braille (1809-1852) in 1821. He lost his eyesight early in childhood though an accident and was then sent off for a special school for the blind. The methods for reading employed back then were extremely tedious, if not to say unusable in practice. Inspired by a system utilized by the french military to communicate in the night, he invented Braille.

Unfortunately, during his lifetime he never saw his system officially used by the institution he worked for (he was a teacher for the blind). Actually, he was forced to teach his system to students more or less in secret. A sad example of patronizing.

Refreshable Braille displays

A few braille cells with reading fingers
  • Portable devices: 8 to 40 cells (characters)
  • Stationary: 60 to 88 cells
  • Usually a multiple of 20 (20, 40, 80)
  • Input:
    • Navigation keys (6 or more)
    • Routing keys (one or two per cell)

These days, all known braille display types have one row of braille cells. I know of just one product from the early 1990s which actually had two rows of 40 cells each.

The length of the row of cells is limited by the type of device. If it should be portable, the maximum amount of cells is more or less 40. Non-portable braille displays can have up to 88 cells per row.

But the size is usually a multiple of 20. Text terminals (DOS, Linux console, ...) very often use a width of 80 characters per line.

Connecting to other devices

  • Serial, USB, Bluetooth
  • Usually relatively simple custom protocol
  • Idea: Implement standard protocol (vt100?)

Current pricing

  • 24 cell display: 3600 EUR
  • 40 cell display: 5500 EUR
  • 40 cell notetaker: 9000 EUR
  • Cost has not changed in the last 10 to 20 years
  • Almost impossible to get in poor countries
  • Or if the social system does not want to pay for it (USA)

The idea

  • Create a specification for all the necessary parts
  • Reuse proofen technology (piezo cells)
  • Keep flexible to allow for different applications
  • History: http://www.vt100.at/?site=braille

Refreshable Braille cells

A refreshable braille cell needs to fullfil certain properties to be useful in practice:

  • It needs to be robust enough to survive (hopefully) at least 10 years of daily use.
  • It needs to be fast enough to acomplish several updates (refreshes) per second.
  • And, implied by the previous item, it needs to have low latency, the user should not be required to way an excessively long time until the dots are all changed to a new position.

This might seem unnecessary to mention latency, but I was indeed once informed about this "revolutionary new technology" to build braille displays really cheap now. The idea sounded pretty and neat until I finally realized that the device would need around 2 seconds to refresh the currently shown dot pattern to something else. No matter how cheap this device would ever be, it would also be next to unusable in real world application.

So while I do not want to prematurely judge new approaches to the problem of refreshable braille, for the purpose of this project, we should make use of the best method currently known, and that is, supprisingly enough, after more then 20 years, still the reverse piezoelectric effect.

The piezo effect

See Wikipedia: Piezo effect for a nice description of the wide applications of this property of certain materials.

Note that we are using the reverse piezoelectric effect to move the pins of a refreshable braille cell. And yes, there is also an application for the direct piezoelectric effect, namely that strain put on the pins generates an electric field. In other words, a braille cell can also detect when it is being touched!

So far there is only one commercial vendor making use of this. It is unclear but they might have filed a patent on this (obvious) feature.

A piezoelectric braille cell

Closeup of a single piezo braille cell

  • Stackable (sometimes in blocks of 8 cells)
  • Usually with one (or two) button(s) per cell (routing keys)
  • Uses shift register (clock and data line) to set individual pins
  • Needs around 250V DC (varies between models?)

Examples of companies which sell piezo cells

Firmware

Open questions

Where do we go from here?