Hackaday Prize Semifinals Video

The Problem

There are roughly 3,600,000 deafblind people in the world, and probably the main issue they face is lack of communication with other people, as they are only able to have a conversation with a person physically right beside them, and knows how to communicate with them. This further limits people with multisensory disabilities to have regular minimum social interaction. Another barrier is their ability to interact with technology, the primary way that they exchange information is by the sense of touch, our devices (e.g. radio, television, computers, tablets and smartphones) do not have haptic feedback, and buying a refreshable braille display could be very expensive (from $3,500 to $15,000).

The Tact-Tiles Approach

What I want to do is to create a API and a scalable hardware platform that can make viable (easy and cheap) the creation of gadgets for people with multiple sensory disabilities, thus, allowing interaction with current devices possible.

The haptic interaction is made by little pads which are touch sensitive and have a vibration motor for feedback, these pads can be arranged in every possible way, making it possible to create various types of gadgets. All the input and output gestures are processed by an easy to use API.

A Tact-Tiles device can utilize up to 32 pads if the developer chooses to use the Touch-X board.

When this project is completed all the information will be available for the construction of custom pads and boards compatible with the API. This will improve the flexibility of this platform for creating new types of gadgets:

The API

The API has abstraction layers for simplifying the work of the developer:

The software handles the input using a finite-state automaton, changing its states accordingly to the data provided by the pads. The automaton is stored in a Serial EEPROM chip using a very organized way:

The Wired Glove

The wired glove is the first application of this new platform called Tact-Tiles. It was constructed accordingly to the rendering below, the pad disposition was defined based on the Malossi alphabet for deafblinds, best described in the page of a similar device, the dbGlove.

Here is the complete schematic for the Touch-X v2.0 board, it was designed using the KiCad suite this way everyone can edit it using free software!

Current Status

The second version of the wired glove has been built at the beginning of this year (2017) and showed a good progress in terms of usability and ergonomics.

The API is under development yet. I still have to organize the source code, but I will try to keep the github page updated.

Hope you like the concept and help to spread the idea! :D

All the software used in this project is licensed by the General Public License 3.0

All the hardware used in this project is licensed by Creative Commons CC-BY-SA 4.0 and Apache 2.0

All the other materials (e.g. texts and images) are also licensed by Creative Commons CC-BY-SA 4.0

Project Logs

Collapse

2017 New prototype
Anderson Antunes • 03/25/2017 at 15:00 • 0 comments

It's been a long time, but I have some good news:
Final Prototype First Run
Anderson Antunes • 09/20/2015 at 22:30 • 0 comments

In this video you can see the assembled prototype running a simple demo test of the touch sensors array:Hope you liked! :D
Prototype Finished!
Anderson Antunes • 09/19/2015 at 18:55 • 0 comments

Finally I finished the prototype, everything is working as expected, apart from the Bluetooth module (I still have to buy a new one). It was the most interesting build that I have done and I am pretty happy with it. Hope you like it:
This night I will be working on the software, so the new video will be up to date. Tomorrow I will post a simple video showing the progress and on Monday I post the new one. Thank you for your support!
Final prototype almost done
Anderson Antunes • 09/19/2015 at 10:40 • 2 comments

I’m only a few hours to finish the final prototype, but I wanted to share the current status with the community. Here are some photos from last night.
I’m still thinking in how I will be able to demonstrate in the new video the vibration of the pads. In the last video if you watch carefully I think you can notice the vibration but on last night’s test I wasn't able to tell which one was vibrating by looking at it, only by feeling and hear it. If you have some idea on how to show this on a visual demonstration please let me know!
16 pads are finished
Anderson Antunes • 09/19/2015 at 10:35 • 0 comments

Here are some pictures of the assembling of 16 pads:
Note the little bumps made for reading the cell letter in Braille:
Making all these pads was a very time consuming task that took me almost five hours to finish. I had no time to make mistakes, luckily just one of the 80 very thin wires broke, but I was able to identify and fix it.

I used 5 minute Epoxy for keeping everything in place. I never had used it before, but the results was satisfactory.
Serial problem with the new board
Anderson Antunes • 09/16/2015 at 10:23 • 0 comments
My test on the Bluetooth module did not work, It can send messages to my cellphone but it always receive “0” when I try to reply them (displayed as a blank line after the letter in the image below).
Oddly I found that every time that I sent a “y” it returned a “@”, this made me realize that it might be a software problem, but when I tried the same code in my Arduino Mega 2560 it worked... So I still don't know what it is all about, but I have some guesses:
- Maybe I overheated the uC during soldering
- Maybe the board has some design error (probably not due the simplicity of the track, see image below)
I also tried to make a serial connection between my Arduino Mega and the board and got the same results.
I think this is a problem with this unity only, so tomorrow I will assemble another board and see if this is really the case.
If someone has an idea about what can it be, please share! ;)
UPDATE: I assembled a second board and got the same results, I also tried making a connection with my Arduino Mega again and it worked! So probably the problem is in Bluetooth module, but is still odd that it works perfectly with my Arduino Mega. Anyway, I bought another Bluetooth module, and I will resume my tests in the weekend.
3rd prototype built
Anderson Antunes • 08/28/2015 at 22:35 • 1 comment

Yesterday I assembled the main board as expected, I used only one slot for the IO, making it capable of handling the first 8 tiles. I also assembled 4 tiles of random Braille letters.
The expected results were achieved, I tested the inputs and outputs using a simple program that reads the inputs and activate the vibration motor of the corresponding tiles.
New boards arrived...
Anderson Antunes • 08/25/2015 at 22:55 • 0 comments

... but my new soldering iron only comes tomorrow, and my flux pen on Thursday, so probably by Friday I'll have some news.
Here are some pictures:
Version 1.0 vs 2.1
Project Back on Track
Anderson Antunes • 08/20/2015 at 19:51 • 0 comments
The main cause that I wasn't able to build a improved prototype yet is the huge mistake of choosing the wrong footprint for the ULM2803. I couldn't find this component at my city so I wasn't able to check if it had the right size, only when it has been delivered from China one month later. So I had to create a new design and order it again, this time using my voucher from OSHPark.
I still had two months left for the construction of the prototype but I had troubles with customs and I couldn't get them on time before I left Brazil. These are some pictures of the boards that arrived in Brazil after my departure:
I'm now in Ireland, starting a scholarship programme in Computer Engineering. I'm loving this country, but until last week I didn't have a address to order things. My boards are now estimated to be delivered on 25 of August, so until there I have to wait.
There are two exciting new improvements in this version of the main board:
- You can populate the board according with the number of pads (input and output) that you will use, e.g. 8, 16, 24 or 32:
- A unusual footprint for the ULM2803 (8 outputs) made by me is used for enabling the use of an ULM2003 (7 outputs) if you want:
New PCB design
Anderson Antunes • 05/27/2015 at 22:20 • 0 comments
I'm currently porting the PCB design for KiCad, increasing the TouchX board hardware openness! :D
2.0 version changelog:
- Added mounting holes
- Added some forgotten capacitors/resistors
- Changed position of the Bluetooth HC-06 module connector
- Added Bluetooth RX pin voltage divider selection jumper
- Fixed SOIC-18 footprint
Here are some 3D renderings of the board:
On the next project log I'll post the full schematics and PCB layout files ;)