Hardware ideas

From OLPC
Revision as of 18:44, 19 December 2007 by Mcfletch (talk | contribs) (General project ideas from the original developer's program page)
Jump to navigation Jump to search
Projects and proposals    +/-
Content ideas Content projects
Hardware ideas Hardware projects
Software ideas Software projects

A list of hardware ideas. See also Category:Hardware ideas

General Project Ideas

  • power management in the device drivers: for us, every joule matters, and a simplistic "oh, we mostly have most of a chip turned off, maybe" isn't good enough. We want to know that every possible power savings has been realized, and that suspend/resume is rock solid and blindingly fast.
  • fast suspend/resume: We must go beyond the current state of the art as discussed at the power management summit.
  • modal operation: if certain applications are full screen, the system should automatically suspend and resume whenver idle for more than a short period.
  • variable speed display driving: (aka: mode change on the fly), again, to save power.
  • wireless: we will be deploying mesh networking. Serious experimentation in this area is in order, to shake down the drivers and to gain experience in its behavior in differing conditions (e.g. rural areas with low noise characteristics; busy metropolitan areas). We understand that to do serious tests, more than a single board will be needed. Please be realistic in your expectations: two boards is not interesting; two hundred boards can't be provided.


Project: The ten dollar e-reader (10ER)

http://wikibox.googlepages.com/home

Experimental idea: smartphone

THE CONTENT BELOW IS JUST AN EXPERIMENT (TEST MOCK-UP), NOT A NEWS ABOUT PILOT MODEL FOR A TECHNICALLY MATURED PRODUCT

We are now making a prototype of XO Smartphone prototype based on Openmoko and Chumby frame as well as various linux smartphones such as E28. Our ultimate purpose is NOT to make a brand new machine, but ONLY to reduce the physical size of XO to hand-held machines.

An OLPC Smartphone is for college students of South Korea, almost 100% of them now having both (laptop or desktop) computers and mobile phones. So, our goal is to provide an XO derivative which can substitute those computers as well as mobile phones. Though the Larger XO for secondary school students can also be applied to college students, we think a mobile-phone type XO will be much more easily adapted by those students, because of their peculiar interests on mobile phones.

The hardware and software architecture is all the same to those of OLPC, except;

  • perhaps MX31 or MX31L rather than x86 processors. As you know, changing CPU of OLPC makes all peripheral circuits to be changed also. Our interests are how to move the SUGAR architecture to a new hardware frames such as Openmoko-based FIC Neo1973 smartphone, which even without marvell mesh wi-fi chips.
  • additional input/output device, consisting of one display and one keyboard.
    • we hope a college student may select any size among 12.1", 13.3", 14.1" and 15.4" wide panels.


There are some problems in attaching marvell wifi chips on boards such as that of FIC Neo1973. If those problems are solved, sample test machines will be distributed no later than 15th, May, 2007 (about 100 units). However, we don't expect those test machines be well harmonized with SUGAR. There must be too many difficulties left before pouring SUGAR on those smartphone-type XO models. --User:Php5

ONCE AGAIN, IT'S JUST A TEST. NOT A REAL MODEL FOR MASS PRODUCTION.

Sugar on Chumby

One of the two XO smartphone mock-ups is to sprinkle SUGAR on Chumby, a radio on open source platform.

Changing the current 266 Mhz MX21 freescale cpu of Chumby with MX31 or MX31L (which provide much better graphic function) may be sufficient to load SUGAR on it and to operate external display devices.

However, the board of Chumby is too big to be a mobile machine. We are looking for somewhat smaller boards for Chumby.

Sugar on Neo1973

From the aspect of size, FIC Neo1973 is an appropriate one to load SUGAR on it. Because the marvell wi-fi chips such as libertas 88w8305,88w8381,88w8385 and 88w8388 series are too big (size) and too hungry (power consumption) to be on the Neo1973 board, the FIC team is currently looking for some smaller ones.

However, as we (OLPC Korea) don't need GSM related modules as XO doesn't, we are now removing those chips from both hardware and software specs of NEO1973. By removing those, there will be sufficient physical space and power supply for the Marvell Libertas 88W8388+88W8015 chips.

Additional feature we expect is Dual User Interface, one Openmoko & the other SUGAR. When the smartphone stands alone (as a phone), it's 2.8" display panel will show Openmoko interface on Linux 2.6.x kernel. However, when connected to an external Input/Output device such as our 12.1" wide display panel and keyboard set we design, then SUGAR will show on that wide screen.

No later than the end of May 2007, we hope to show our prototype mock-up sample.

See also Belkin with Marvell 88W8385 integrated 802.11g wireless MAC/BB and 88W8015 RF transceiver devices.

Magnifying glass / macro-lens bugbox

A minor modification to the plastic around the camera would leave a slide-in connector for a magnifying glass / slide microscope / bugbox / macro lens. I'm not an expert, but even just a cheap plastic bugbox dealie, cost of manufacture in the $.10 range, would open up a lot of exploration - a macro lens with a field of view of a few centimeters. Leaving the connector now would let you design the lens later. Would a true microscope be possible at a reasonable price? Note that this would be used with the screen rotated facing up, natural light from above. Homunq 00:57, 28 July 2007 (EDT)

Obviously this idea exists in prototype, quoted from News:
5. $1 video microscope: A video of Mary Lou's prototype microscope attachment for the XO video camera is posted on the web (Please see http://www.youtube.com/watch?v=wI28-IS9AII). In the video, she compares various LCD screens. The microscope, which has ~ 100× magnification, could be useful to analyzing water quality, among other things.
My question is, how is this connected to the laptop? Is the connection hard to build or flimsy? Homunq 16:23, 8 August 2007 (EDT)

display swizzling

The swizzling is not optimal for all tasks. It would be good to let software load a 3x3 or 5x5 convolution matrix to adjust things better. Currently you have:

 0  1  0
 1  4  1
 0  1  0

(divide by 8)

A much better matrix for eliminating color artifacts is:

 2  1  0
 1  4  1
 0  1  2

(divide by 12; scale everything up if you wish to approximate with a shift)

In some cases the following could be more attractive:

 2  1  1
 1  2  1
 1  1  2

(divide by 12; scale everything up if you wish to approximate with a shift)

For apps that wish to do sub-pixel rendering, swizzling needs to be disabled. Note that this is distinct from greyscale mode, which needs to mix red/blue/green in the proper proportions for human vision. Disabled looks like this:

 0  0  0
 0  1  0
 0  0  0

(divide by 1)

BTW, hopefully the swizzling is done after the gamma look-up table has converted the data to linear. Failure to do so would be a cause of some nasty artifacts.

TeleHealth Module

The idea is to connect electronic instruments plus sound and video to a wireless-equipped low-power computer such as the OLPC Laptop, and send data back and forth between a village and a doctor's office in a town, thus multiplying the effectiveness of doctors. It would also save villagers much time and money compared with taking busses or walking to town, or worse, not being able to get there at all.

The result is that a health worker with modest training can carry out most of a standard medical office examination, to be interpreted by a doctor in real time. The doctor and patient would be able to see each other and talk to each other.

There are some limitations on this process. In particular, the doctor would be unable to touch the patient and feel for abnormalities. But the doctor could decide which patients needed to come in for further tests and examination, and which could safely remain at home and be treated there.

For an overview of the technical ideas, see TeleHealth Module

Keyboard

For the sight impaired
A Braille embossed, transparent(silicone rubber), and removable Keyboard overlay.

B-1 Keyboard layouts

default QWERTY Layout
Thailand Layout


Further information can be browsed checking the Category:Keyboard and Keyboard layouts; or more general, Category:Hardware, Category:Language support or Category:Languages (international)

For children

Why stick to QWERTY layout? Maybe a better layout can be applied here... a simple sequential ABC... layout would let the child learn faster where the keys are. Remember... Most of those kids would have never seen a keyboard before. I like the idea of "removable keyboard overlays". that idea could be applied here, once the kid is familiar with the keyboard function and usability, he could overlay a more standard "QWERTY" layout and start learning to type in such. Or maybe this will be the end of the QWERTY layout for good :-)

Preliminary partners

Summary

OLPC Korea shall make a few voip phone prototypes; one of smartphones is based on neo 1973 and Belkin, and one of wifi phones is based on VM1188T phone. This product will be distributed after 1~2 months later deploying XO laptops in Korea.

Candidates

Partnership information shall be disclosed after signing agreements between participant companies and our joint venture.

Additional power

The XO solar power charger raises a number of interesting thoughts. The random thought paths that come off this include everything from solar calculators to windmill-powering houses. These lead to more sane and directly interesting thoughts.

Solar laptop case

The XO can charge via a solar power adapter. Why not, then, take the next logical step and build the adapter into the laptop? Consider the entire back of the screen having a hard, transparent plastic coating; underneath this lies a solar panel. Whether closed or open, the solar panel still collects light and charges the battery.

A number of interesting considerations stem from this:

  • Use the solar panel continuously
    • AC power (i.e. a second solar panel, or human power) should supplement the built-in solar when present
  • Deploy a backpack with a soft plastic window, such that the placement of the XO can expose the solar panel to the window and the light outside, allowing charging on the way to school or anywhere.
  • Can we make the case removable, sealed, locked (latched, not keyed), and still thin, such that the clear plastic cover comes off allowing the replacement of damaged solar panels with newer, possibly even more efficient models?

Inductive charge stations

Here I assume the students go to some sort of school or the like to get educated, rather than self-tutoring since age 3. In this scenario, we can envision an inductive charging station using no wires or electrical contacts.

For those unfamiliar, specific types of transformers adjust voltage and current via using an inductive transfer of electricity. Basically, two coils of wire don't touch, but they sit close together (even interleaved) and the electromagnetic field induces power from the hot one into the cold one. Their different construction (i.e. number of twists per cm) cause a difference in voltage and current.

More interestingly, certain products such as some electric toothbrushes use the same design to actually charge a battery. The battery sits in the toothbrush handle, a sealed unit with no electrical contacts. When the user places the toothbrush into a base station, a coil at the base of the toothbrush sits inside a coil at the base station, which induces power into the toothbrush and charges the battery.

For our purposes, the bottom of the XO laptop could contain a coil for inductive charging. Placing the laptop upon a surface with an inductive charger underneath would transfer power to the laptop. Power could come from wind or solar power on the actual building (such as a school roof), stored into batteries. The students' desks would become the charge station, with the desk surface as the charge surface.

Unfortunately this wastes power. Induction doesn't give the best efficiency in power transfer; and having the inductor on all the time also uses power. One solution could involve using a much lower power sensor to detect a laptop, for example a weak inductance charge that drops when you place a load across it (bringing the laptop too close does cause a load); with this scheme, the inductor could increase power when it detects the presence of a laptop, reducing waste. Another simple solution involves just putting a switch on the desk and turning it off when not in use.

Optimally, using both prescribed methods will reduce power when the laptop has a nearly full charge or when the student leaves his desk. Resistance in the charge circuit prevents inductance into the laptop, I think, so make the amount of power adjust proportional to the load (more load, more power) and the inductor will use as little power as it can. The desk will supply minimal power with minimal load (i.e. no laptop) and thus automatically behave as in the first prescribed method. To extend this, have the desk use a solenoid to physically flip the switch off after maybe 10 minutes like this; then the power automatically gets cut when not in use for extended periods.

Combining this kind of base station power supply with the internal solar power described above and the external human/solar power already present in the current design gives a largely efficient collection base for power. Other interesting ideas include making the surface of the desk a solar panel to charge the laptop and reduce load on the battery store in the building (or the power grid for the more fortunate), in case there's light inside the school where this happens (this also encourages skylights though).

This of course means the OLPC organization would have to design the inductance-based charge surfaces or desks too....