Tactile braille displays?

Orbit Research has the Graphiti and Graphiti Plus. These are commercially available. The Graphiti has a 60x4 dot grid. The Graphiti Plus has the same grid plus a 40 cell display. Unfortunately while Orbit has promised the Graphiti Plus will be able to display images in real time sent to it over HDMI same as a monitor, this has not been released yet. So you have to manually save images to the device such as by using an sd card. It is doable but tedious.
I think the older Graphiti may have working HDMI but if you need ral time access like a monitor you may want to ask.
The other one is from Dot, the DotPad. (Dot is the company that makes the Dot Watch the Braille smart watch.) However the best I can tell it isn't commercially available. A few lucky people have them but it seems you can't just put in an order and get one.
APH is also working on a tactile graphics display. This is also still in prototype and we can't order it either.

I think it is. As long as money is no object.

If memory serves Monarch is the name of APH's project/device.... It could have the most potential of all since it can also display multiple lines of Braille, but first they need to ship it.

Here:

https://www.aph.org/soaring-to-new-heights-a-professionals-perspective-on-the-monarch/

I understand there are units around for testing. According to the article you linked too it says, "The Monarch will be publicly available in September and eligible for purchase with Federal Quota funds. Stay tuned to the website, APH News, and your email
inbox for more information on this upcoming device."
Which is more current than any information I had before, thanks. I still would not call that shipping now though.

I just read the date, yesterday. I should have gotten AI to summerise! You would think I would know better.

This device seem to do at least part of what you (and I) are both looking for. Hoowever its status is not very clear, it was available back then but for quite awhile we didn't have any price displaied but when I checked three days ago it seems to be available again. It's way more affordable than any of the solutions / coming prototypes mentioned above too, although I just read the documentation and didn't bother asking for possibilities of demo.

Travis Roth already mentioned the dot pads I know about, but I'll make a few comments. I work in accessibility R&D, and I've read about the history of tactile graphics, created my own tactile maps by hand, talked to people trying to develop large area graphics pads, and so on.

Many engineers have known about the desire and the need for large tacticle displays, but the economics are tough. The devices are tricky to manufacture. As you know, assistive technology tends to be wildly expensive, and tactile displays are no exception. There are only about 100,000 to 150,000 people in the U.S. who know Braille, and only a small fraction of those people would ever be reached. And to top it off, far more sighted people talk about the need for accessibility and applaud efforts to improve accessibility, but then don't do anything to get properly engaged.

I should mention that I'm sighted, but I have some Braille typing skills, and when I'm in practice--which isn't true today--I do an okay job reading Braille slowly by sight.

The Graphiti is probably the best-known device, but from what I've heard it's very expensive. A friend of mine who tests just about every new bit of tech that comes out wasn't impressed by the device, but I still think it has potential.

Alex Russomano and the team at NewHaptics in Michigan have been working on an alternative to standard mechanical Braille cells for some time. They attended CSUN earlier this year, but I wasn't able to go, unfortunately. Here's the website for NewHaptics: NewHaptics.

When I spent some time in the Dot Corporation booth at CSUN in 2020 I was impressed by their Braille cells. The cells are very light, and they click together like Lego bricks. Potentially, a Braille or graphics display of just about any size could be created with their technology.

I have a Dot Watch, which has just four cells. When I showed the watch to friends reviews were mixed: reading the Braille can sometimes press one of the dots down. The cells are covered by a thin, flexible layer of plastic to keep dirt from getting in, but the plastic also affects the feel of the cells.

There are haptic devices that could be used to create tactile graphics over arbitrarily large areas, but those haptic devices tend to be very expensive.

A few years ago I purchased and tested a TanvasTouch haptic touch screen. It's cool, but it was expensive. The design would have to be modified to be properly useful. If I can dig up the TanvasTouch screen and get it running again, I may create a Medium post or a YouTube video about it.

According to Dot Inc, the Dot Pad uses new and improved cells than the Dot Watch has. I wish they'd make a new Dot Watch with the new cells. Anyway it seems we're so close but yet so far away.

Now in 2025, any improvements or user feedback from the us especially with monarch? Also Apple say VoiceOver can handle multiline braille display with 0 additional detail, there are already weird bugs with regular braille, has anybody been able to test monarch with an apple device?
I will be in CS soon and if money permits, I am in Quebec, could seriously think at least ask around about how to buy this, the first blind equivalent of a car :) but if it can help why not...

Monarch is available now. Though I don't know anyone who has ordered one. Anyone who can spot me $18,000 I'll report back! Interestingly, Vispero has a blog post up recently saying they're working on multiline Braille support and they mentioned four partners including APH and Dot Inc. This is the other piece of the puzzle. Orbit Research has the Slate 3 and 5 line options out now, but no screen reader uses them effectively.

Yeah. Everything considered I think for my intended use case I'd rather buy a great braille printer instead...

Here's a link to another company doing multiline displays in a modular way: https://www.tactile-engineering.com/cadence

Actually for my future needs in the next years I'd be probably getting both a good high quality embosser with a high numeral cost :) and piaf from humanware, I have experience with both outputs and each has advantages the other doesn't.

I'm not sure what you mean by piaf from HumanWare, can you clarify?
As for Braille printers, you will get much more resolution, even the basic ViewPlus I have does 100 dpi. But, very slowly.
This leads to another item we need for either type of display: more graphics preprocessing to make it tactile friendly. For example, make braille numbers, reduce unnecessary artifacts, improve contrast for the important details. Often graphics for the human eye are just too complex for tactile. I hope AI can help. There are image generators already, so I hope someone with more skill than me is researching it.

This leads to another item we need for either type of display: more graphics preprocessing to make it tactile friendly. For example, make braille numbers, reduce unnecessary artifacts, improve contrast for the important details. Often graphics for the human eye are just too complex for tactile. I hope AI can help. There are image generators already, so I hope someone with more skill than me is researching it.

I actually have some plans to work on this kind of thing once I'm officially out of my current job, but won't provide details yet. The translation of Arabic digits to Braille wasn't something that I was considering but that can be easily achieved as well, and I think it's a good idea even though I don't use Braille myself. Graphics embossers aren't my end-goal, but I'll be using one myself during development so I can make the GPU shaders available in a way that everyone can consume if I find that they offer a good enough value for general-purpose use in my tests.

Edge detection is theoretically quite easy to do algorithmically by creating a feature map containing the cosine differences of normalized chroma vectors multiplied by luminance, and then filtering out anything below a certain threshold to produce a final image where all the edges are turned into outlines. The only challenge here is figuring out the right threshold and eliminate small high contrast details like dark spots on people's skin that aren't really useful information.