A Raspberry Pi For Audrey, Perhaps?

The local Makerspace is discussing the new Raspberry Pi, a $35 PC with a credit card sized form factor.

As an Arduino user, $35 is a nice price point (I can buy an for about that price), but it’s no good if it doesn’t do what I need.

It does.

In spades.

It turns out that the device outputs composite (RCA TV) or HDMI video signals, can send audio to speakers or headphone, and has both a network (RJ45) connection and USB ports. It gets its power from a USB power jack, and it even includes an SD card slot for its ‘hard drive’ – although with a USB hub you can connect all kinds of things to it, like an external drive, keyboard, mouse, and so on.

As for power, it’s comparable to a Pentium 2 at 300mhz, but graphics are at the Xbox 1 graphics level. Finally, it has 512meg RAM memory built in.

Of course, for Windows, this all might sound somewhat anemic, but running a version of Linux from the SD card, the computer is very, very nice. It’s goal is to make programming accessible to school children again by giving them something ‘closer to the metal’ to work with, but frankly, I think the foundation will have a huge hobbyist market clamoring for it as well – I know I want mine ASAP (or sooner).

However, aside from a neat gadget to play with, it has very real possibilities – imagine:

  • CNC – the computer has a GPIO (General Purpose Input Output) port, which can bring out various signals. While I haven’t found out too much so far, it seems to be all 3.3v logic levels. If enough pins are available for hardware, I can envision a CNC machine connected to this (or actually, it IS the CNC machine, with the pins simply run out to sensors and the 3 motor drivers of a CNC machine). Currently, the CNC design process is to create a file and send it to the host computer, which chops it up and transmits it to the CNC machine. With this device, chopping and cutting can all be done directly on it. Send it the file via Ethernet, and have it manage the conversion to G-code, AND do the cutting. Connect a monitor, keyboard and mouse, and you don’t need an additional computer to run anything!
  • The Audrey Braille device. My main reason currently for watching this. Using an Arduino Mega and motor driver board, I an spending close to $50 to drive two motors, and communicate with a computer via serial USB. Using this device, it IS the computer, AND the motor driver, AND anything else I need! As long as I can connect two motor drivers to the GPIO (about 8 I/O pins) as well as a switch multiplexer (three more pins), I should be able to switch over. Now, instead of connecting to a laptop to run, the device can do it all. And the extra audio capabilities are perfect for the Blind – as Linux boots up, they can get audio diagnostic messages. Plus, with built-in network connectivity, and perky performance under Linux’s Text mode (which is actually better for the Blind), this truly could be all a Blind person needs for Internet and computer connectivity (and Braille display), all in one package.
  • Projects galore. The GPIO brings out signals for SPI, I2C, and the UART, which opens up a lot of communication options. I see this device as the ‘glue’ for a whole host of projects, from devices that hook into the Ethernet (cheap media server, anyone?) to the basis for cluster computing (you can connect thirty for the price of a decent computer!), to one-off hobbyist projects.

I’m hoping in the next few months the backlog on ordering will end, and I can get one. Frankly, I’m wondering if I will continue with the Arduino if this gets popular – after all, similar I/O but in a Linux/PC format is a powerful combination for projects – and the price is right, too. In fact, at $25, you can get it w/o Ethernet, so the price gets even better for specific hobbyist projects.

If I had any comments, it would be for two small wishes: Power through Ethernet (which is going to be supported eventually), which means this device wouldn’t need a wall wart; and a built in real time clock (which is too expensive to add to the $35 model). Of course, the software can poll a time service over the network, so for many situations that may not be an issue.

Gripes aside, it’s a fascinating project, and I’m looking forward to playing with one real soon…

6 thoughts on “A Raspberry Pi For Audrey, Perhaps?

    • Yes, I’d just read that Make article, and wondered if the final code could fit in the Tiny85’s 8k – if so, the chip is a really neat idea. Thanks also for the USB article link.

  1. You’re pretty far along with this design, but I thought I’d mention something from U.S. patent 4752772 (issued 1988, http://www.google.com/patents/US4752772?printsec=drawing ) points out in its summary: “… currently available Braille display terminals limit user efficiency and, in an effort to eliminate the limitations, we have discovered that efficient and effective Braille reading does not require a multicell display. Comparable reading speeds and comprehension can be achieved with a single cell display which the user can electronically scan over the target text.” They then describe a six solenoid braille cell built into one key of a computer keyboard.

    • That’s really neat – I’d be curious as to how a blind persons reacts to a single character like that – certainly, part counts would be very low!

      • The DEC researchers say in the patent (which by the way, is now in the public domain as it’s more than 17 years old) that “Reading speeds at the single cell display are comparable to the reading speeds acheivable at the multicell display; but the wasted movements associated with shifting the hands back and forth between the keyboard and a separate reading display are eliminated. Like the sighted user, the visually impaired person can type, verify, read and edit while keeping the hands in the same location. This results in a marked improvement in editing efficiency. The immediate feedback which the user receives through the “j” key leads to improved error rates, quicker discovery of typing errors, and less serious errors. Thus, with the key-embedded Braille display, convenience is enhanced and user productivity increases.”
        Another discussion of braille reading speeds at http://www.braille.org/papers/invea/invea.html found that that “In another condition of Troxell’s experiment, experienced braille readers read text presented a character at a time by sensing patterns of pins pressed against the tips of the fingers … his subjects were reading not braille characters, but analogues of braille characters. Nevertheless, after only brief practice, they achieved a mean reading rate of 18 wpm, which was the maximum rate at which the instrument used by Troxell to display characters could be operated.”
        This newsletter from NASA summarizes what pin sizes, spacing, and force are needed for best perception: http://ndeaa.jpl.nasa.gov/nasa-nde/newsltr/WW-EAP_Newsletter11-1.pdf#page=6
        That newsletter is from the electroactive polymer research group; it’s another way to create haptic feedback devices.

        Direct electrical stimulation can also simulate mechanical pressure; section 4.1 of this SmartTouch paper by Kajimoto et. al describes the electrical parameters that best simulate mechanical pressure: http://files.tachilab.org/intconf2000/kajimoto200303HS.pdf#page=5

        The original objection to solenoids (power, size, expense) might be alleviated by using Plantronics parts; they sell some very compact, low power solenoids: http://www.microflight.com/Online-Catalog/Actuators-and-Servos

  2. A 1 cell display would be very inefficient. You can’t do reasonable editing of any sort with that model. There is a reason that that model was not adopted I think.