Refreshable Braille, within reach.

Most Braille readers never own a refreshable display. This cell swaps the piezo stack for a printed cam and magnet.

A finished Braille module held between finger and thumb against a dark background.

Grand Prize, 2023 Hackaday Prize

The refreshable Braille module

Only about one in ten blind people reads Braille. The single biggest reason is that the hardware costs more than most families will ever have.

A commercial refreshable Braille display prices each cell at $100 to $150, because each dot is driven by a piezo bimorph. A forty cell display is therefore a car.

This module replaces the piezo stack with an eccentric cam carrying a 1mm by 0.5mm rare-earth micromagnet. A small coil flips polarity, the cam rotates between two positions, and the pin rises or drops. Once it is up, the geometry holds it there: the weight of a reading finger cannot back drive it, so a static page draws no power at all.

The housings are SLA printed. The coils are 50 micron enamelled copper wound on ferrite cores with a winding tool that is itself 3D printed. Peak draw is 1W while a single pin actuates, and adding ferrite cores took the refresh from 200ms per pin down to 50ms.

Raw materials come to roughly one dollar a cell. It won the Grand Prize in November 2023, the final year of the Hackaday Prize's ten-year run. The project is open, and it was built with a team credited on its project page.

Three Braille cell circuit boards, each about the size of a fingernail, resting in an open palm.
Scale

Three cells, one hand

The boards that carry the coils. Each one is a single Braille cell.

An assembled Braille module on its circuit board, held between finger and thumb.
Assembled

One module

Housing, coils and board together.

A hand holding an SLA build plate covered in freshly printed resin Braille cell housings, still on their supports.
Printed

Straight off the build plate

Resin housings on supports, before cleaning.

Dozens of small printed Braille mechanism parts scattered on an orange cutting mat.
Mechanism

The parts that move

Cams, followers and cell bodies, mid-revision.

How it works.

Four parts per dot. The cam carries the magnet, the electromagnet turns the cam, and the geometry holds the pin where it was put.

The assembly, taken apart and put back together. Labelled on the drawing: PCB, electromagnet, cam and magnet, Braille dot.
A single micromagnet resting on a fingertip, arrowed, barely larger than a grain of sand.
The part the whole cost argument rests on. One millimetre across, half a millimetre tall, and there is one of these behind every dot. A piezo stack does the same job for a hundred times the money.

And it moves.

One module, cycling. Each dot is a printed cam and a micromagnet.
Three modules on the driver board, under a microcontroller.
The coil winder. It is itself 3D printed, because 50 micron wire is not something you wind by hand.

Ten years on one problem

The Grand Prize was the seventh year of work, not the first week of it.

The first version was published in 2016, with Paul D'souza. It used vibration motors from mobile phones, with the eccentric weights replaced by shaped cams, to push the pins of a printed cell up and down. Hackaday covered it that April.

It worked, and it was still too expensive and too slow. So the mechanism was thrown away and redesigned around micromagnets, which is the version that won in 2023. Seven years between the two, most of it spent on the parts nobody photographs.

A white 3D printed Braille cell held between finger and thumb, its raised dots visible against the plastic.
2016

The first cell

Printed, pinned, and small enough to hold between two fingers.

A printed Braille cell lying beside a two rupee coin, showing that the whole cell is barely longer than the coin is wide.
Scale

Next to a two rupee coin

A whole cell, barely longer than the coin is wide.

A finished Braille module held between finger and thumb against a dark background.
2023

The module that won

Same size, different mechanism, and no piezo anywhere in it.

The PCB artwork for the Braille cell evaluation board, red and blue traces on black, seven cells across the bottom edge.
Driver

Eight cells on one board

The evaluation board, which is where a mechanism stops being a demo.

A pencil sketch of the Braille mechanism, labelled with followers and raised portions, beside a hand-written gray code table.
Before any of it. Followers, raised portions and a gray code table, worked out in pencil.

What other people wrote about it.

Unlike commercial units, which use expensive piezo-electric actuated pins, his design utilizes an array of tiny electromagnets to rotate even tinier cams.

HackadayHackaday Prize 2023: Ending 10 Years On A High NoteTom Nardi, November 2023

Only about 10% of blind people around the world can read Braille. One primary reason is the high cost of Braille displays.

HackadayRefreshable Braille Display And Braille KeyboardApril 2016

Tell me what you are building.