Most games work with visual clues, which results in an decreased fun for visual impaired people. As a sidetrack of the Personal Photonics project (https://hci.rwth-aachen.de/pp), funded by the german ministry of education and science for developing an haptic toolkit, these game modifications happens due to inspirations of the community. They are still mostly first steps at the moment, just to check if these ideas are feasible and make fun.
Results so far: From a usefulness perspective at the moment the big buttons are the most ready parts (driving around the world on a wheelchair in use), but have little to do with the intend of "haptic" games due to the original personal photonics project. The scorecard gets also good feedback, but the last version still needs testing on the long run if it works as intended.
The water gun might become interesting, at the moment we are using (and built) them for ourself, might be interesting if there is an actual more "serious" use case... as serious as one can get with a water gun :)
While pong on the belt works nice, buyilding it just for ithis purpose is overkill (there the 1D-games might be easier to built and use.
The construction game pieces fail due to stability issues or (when cirmcumventing these throug bigger build size) space constraints, while memory is not as funny as intended (as one could have already known beforehand).
From all the ideas so far the scorecards are the most useful thing (both from the usage as well as production-wise ^^). The biggest problems there is the (minor) problem of the magnets flipping over to another line, which can be reduced by using deeper holes or bigger distances between the checkboxes,abacus rows. The latter will make the card more clumsy, thats why we will go for deeper holes, which will also be used for clearer distingushing of the abacus positions.
While the washer systems gave overall a great locking feature, when the magnets clicks into its hole, some of the washers glueing failed over time, as well as the production process of gluing washers in would be rather expensive (but with reusing common materiels it's of course great for single rebuilding processes as an open-source project). Instead we are now testign the use of a single sheet of metall behind as a base plate. The 3D- cover is then glued on top of it as a single piece, with its form allows even to use the most basic injection molding, thermoforming or casting.
This will furthermore elevetas the issue with the not-fingertip-friendly 3D-printed braille letters, since in this case the right half-sphere form can be used for each dot. For a 3D-printed version it's better to leave the braille letters away and later glue common embossed braille tape with the text on the cards.
Which one need just a (magnetic) sheet (183,7*140) mm, either a magnetic foil or just metal, then a bunch of 3mm magnet spheres and the 3D-printed (or otherwise manufactured) form, glued on top of the magnetic sheet and you are done.
... sort of :) Here we got inspired from the prosthetic Nerf gun (There exist also the idea of a mind controlled one .... hmm, looking at the mind flex controller ...). And since the CCCamp this year was a bit dusty, we considered that the world needs also more inclusive water guns:
Using somthing like the EMG sensor (but basically anything else will work) as a trigger on an analogue input an Arduino Nano and a relais connected to a digital output to drive 12V with a "little bit" of current (delivered by 8 AA batteries in row in two 4 AA batterie housings) to control a wiper pump. It will suck the water out of a bottle and soak adversaries through a nozzle at a wristband. The pump is attached with a bunch of hot glue on the bottle cap, otherwise 3D-printed housings and textile strips (reusing lanyard and similar things) are used to attach the parts on belt (Batterie pack with arduino/relais below and the bottle) and arm (EMG sensor and noozle - but places might vary depending of personal preferences/needs). The also provide a bit of water spray cover, otherwise the good old hot glue is used for encapsulation.
At the beginning we thought that we also need a valve at the bottle to let air in, but at the and, as long as only short pulses are sprayed, the water/air will be sucked back through the pump in between shoots, even preventing/reducing leakage of water through the noozle while not shooting.
Ok, still a little bit far fetched from the original haptic games idea, which turned out to go more into a games for impaired direction), but actually there exist a paper wher there used water pressure in a sort of old crt-monitor-way as an haptic feedback device ^^.
Another interestign approch are 1D-games, which can be played on a led strip. Examples are the 1D dungeaon crawler, pong versions, racing gameeven a two player combat game. Since the output space is reduced, it makes them also more feasible to be used with less controls, therefore our few buttons which can be used with the head are enough (and we have as additional benefit nice demos for them for maker faires :). But that gives another interesting idea: we can also use a single row vibration belt as game field, going away from our overkill HaptiVest to the more common and easier to built 1D navigation belts. Another interesting approach would be (going away form the limitations of the vibration motor things) using a braille line for this type of games. Of course, the player wouldn't have complete overview over the game area and needs to scan the line with the fingertips, but giving a more precise and informative overview over the game area. E.g. the start of the braille line is the player position, and scanning the line would be the search for upcoming obstacles (dungeon), the ball (pong) or the street layput (racing game), to react at the right time.
For playing games with the big buttons, they have to be connected to the computer instead of the wheechair. Here we could be lazy: As described here, one can just use a Arduino Micro or Leonardo as a joystick or gamepad (or just an ordinary keypad). So we just need to make a nice housing, add mono phone jackets for each button and we would be done. But this leaves room for more ideas: Since we can use basically every type of sensor as an input, we might be able to go away from the buttons and add additional sensor types, since game control with just a very few buttons is limited. Simple options are: A gyroscope as kind of joystick (analogue input), and/or air flow sensors as additional input control... But in this case we are in the "its start to get complicated", so that will be thing which is still ongoing in development.
Originally intended fo another purpose, stearing a wheelchair with just a ring of buttons behind the head, we designed big buttons, based upon a standard small button in a 3D-printed housing. Theey should be relative cheap, reliable and needed to be activated with a slight touch. As it turns out, they can be also used for playing games. Not as a arcade button, which would be also possible, but for the wheelchair driver. In this case, a version with a slightly harder button, but a better click-sound for activation was made... which leads to other ideas. But first: Instructions can be found here.
Using the HaptiVision Belt as base we tried out if we can use it also as Display for playing games on it. Of course, with 16*8 pixel that's not high resolution, but we thought that at least games like Pong might work. Adaption was rather easy: Using an Arduino as controller, we added an 16*8 WS2812 LED matrix at pin6 as a display and two linear variable resistors as controllers at pin A0 and A1. The haptic matrix was then connected with the I2C port. So writing the code was for us the biggest issue (not for me, since Moritz Messerschmidt did the work), since we had such a belt already at hand. The Code can be now also found at the belts github page.
Another idea was transalting gamecards for a construction game into sculptures and reliefs - translating the image what to build into an haptic senseable object. Construction was of course rather easy, but both versions didn't work well in real life: The sculptures broke easily, and the reliefs were hard to read, since details where to small. Of course, by printing them bigger you can circumvent these issues, but then you need a very big box for the cards. Design files for OpenScad can be found here.
At the moment the most useful (and even easy-to build) system are the Yatzee scorecard. You just need to print the openscad scorecard file (modify the text corresponding to your languange or comment this part out and use embossed sticker instead for better haptics). Afterwarts glue a sheet of magnetic material behind and add the magnets. And you are already done and ready to play. Really lowtech but useful.