Kindle 2 ($360)	Kindle (discont’d)	Sony Reader PRS-505  ($299)

My major complaints about the original Kindle were that it was flimsy, thick, clunky, and the buttons were not well designed. The plastic it was designed from made its weight distribution funny – so it actually even aggravated my tendonitis. I liked the WhisperNet feature, and the keyboard – but the slowness of the screen made annotating books a pain at best, and referencing the annotations wasn’t really useful enough to merit the design flaws. Amazon’s closed ebook format isn’t great either – most of stuff I want to read just happens not to be available in Kindle format (i.e. academic papers, textbooks, papers that I’m reviewing/editing), so I ended up with the Sony Reader which supports viewing of native PDFs as images, with additional support for portrait or landscape viewing (I wish there was a button), and a zoom button for magnifying the text if you have OCR’d text accompanying the image. Since I didn’t want to take a suitcase full of books with me to Uganda, I sliced the bindings off of them, scanned them to pdf and OCR’d them, and I’m reading them on my reader instead. Much better carrying a slim e-book reader on the plane than the 2-inch thick copy of James Scott’s Seeing Like a State.

The Kindle 2 is even slimmer than the Sony Reader, also comes with a leather case, and has the advantage of incorporating wireless and a keyboards for just $60 more. Kindle has access to a larger selection of copyrighted e-books, “Kindle Editions” at much better prices, with a much cleaner interface. Sony’s software, frankly, is flaky, slow, and crashes a lot. But at least I can put my PDFs on it directly. I think for the Kindle I still would have to email my PDFs to amazon and pay them 10-15 cents to upload them to the Kindle in some weird, potentially mangled, format. I might be able to put up with that from Berkeley, but depending on email access for giant PDFs out here is totally impossible. And I like my WYSIWYG PDFs. In that sort of vein – Sony’s ebook philosophies are actually more “free thinking” than Amazon’s (for all those copyleft people out there) and the Sony Reader supports the open eBook format (ePub), and as Wired notes, actually provides access to more public domain books than Amazon offers on the Kindle through a recent deal with Google Books.

What I really want is for Amazon to build a Kindle that supports GSM, so I can stick a Ugandan SIM card in it and download Kindle books over the local network? Please? Or I guess I can wait until i get back next year…

I have this vague theory that the Kindle devices might make decent computing platforms for rural areas.  Imagine – data connectivity, low-powered devices that don’t need to be charged more than once every two weeks or so, built-in keyboards, screens that are visible in sunlight, a large screen, and a price point comparable to smartphones, or less?  What the heck am I doing working with smartphones with tiny screens, batteries that die in a day, and keyboards that are too small for healthworkers to read?  Oh yeah… waiting for the ebooks to take off, the platform to stabilize and open up, and um.. trying to finish up my dissertation before starting another project. But if amazon is willing to throw some summer interns at me this year, and a few Kindles, I think I could manage to host them here in Uganda.  Umm.  We just need to find an EVDO network or get Sprint to subsidize the roaming charges? =)  Anyone know someone at Amazon?

There’s also a Sony Reader PRS-700, which I haven’t seen, retailing for $399. It includes an LED light (which I think is great, since I can’t get my booklights to attach, and I think it is lame to have to wear my headlamp to bed, or to have to lift the mosquito net to turn off the lamp on my nightstand), and a touchscreen. I can’t imagine how the touchscreen actually works – I think it’s something I’ll have to see to really understand/evaluate.  But ultimately I think Sony will have to move towards integrating wireless into their readers…

Tweet This Post Facebook

The fact was that I spent a lot of time on airplanes and figured it might be worth the investment to spend some money on a headset such that I might be able to hear music and my recorded interview data over the sound of the airplane engines! And sometimes it is nice to be able to sit in a shared office and have some privacy.

The long and short of my research: Noise Canceling technology requires batteries and external power, and is ultimately bulkier and heavier. Noise isolation is great because you can actually turn down the volume being sent towards your ears, they work like earplugs, and it doesn’t require additional power. And they are cheaper. But it can be a little weird because – you are totally deaf to the outside world (i.e. you can’t tell if a flight attendant is trying to get your attention), and if you eat something you can hear yourself chewing, and sometimes they don’t work so well if you are running. Mine are fine as long as the cords are hanging relatively free, since V-Moda redesigned the cables with a cloth cover.

You can plan ahead, but sometimes you have to improvise a bit to make high-tech things work once you leave home...

The V-Moda ones are especially great (not that I’ve actually compared them) because they come with a carrying case. The ones I got came with a little compact leather case that open when you pinch it, and a v-shaped cord wrapper, which was initially a little mysterious to me but now I think is the best thing ever. I want one for all of my corded accessories! Unfortunately the new ones don’t come with the v-shaped thing – they come instead with a rectangular carrying case sized for the iPhone and the iPod Touch. I don’t blame them, but as an iPod Nano owner I’m glad I got my headset earlier.

Okay so that was a long intro. Onto my sob story. I was walking home the other day from work, and I decided to listen to one of my interviews on the way, so I pulled out my Nano, and my headset. As I connect the headset to my nano, I notice that one of the silicon ear fittings is missing. I should note that this happens to me almost every plane flight – these things fall off all the time and it drives me nuts. Why I don’t have spares is beyond me, but they give you three pairs in different sizes (small, medium, large), so I just picked my size and left the other two pairs at home when I came. On a plane, this is mostly fine – I mean, where can a little black rubber thing go on a plane anyways? I’ve always found it eventually. But this time, when i pulled the headset out, I managed to drop the earfitting onto an area of the ground littered with black pebbles, bits of trash, and random grass growth… and a ditch. Yikes. Ever hopeful, I searched anyways, to no avail.

So I walked home, listening to the interview, with one ear happily isolated from noise, and the other ear listening to the interview and the caws of the maribou vulture-storks, with a piece of metal uncomfortably perched in the canal. I had thought through this situation before, and I plotted, considering my options, and wondering if I could wait for someone to bring a batch of spare fittings to me in April…

My solution: well, since noise-isolating headsets are essentially earplugs with sound coming through them, why not use earplugs? So I got out an earplug, my handy-dandy leatherman, and hollowed out the center. The leatherman, as it turns out isn’t all that useful for hollowing out the centers of things, so once I made divots on either end of the earplug, and cut it to the appropriate length, I used the pen from my Palm Centro to poke a hole through the center, such that the middle could grip well onto the headset. I think it actually works better than the silicon fitting – at least it is less likely to fall off, even though it doesn’t look nearly as slick. And now it is much easier to tell the left from the right.

It goes to say that 1) I’m pretty privileged to have the gadgets that I do have, 2) to have the options that I have with which to fix them and 3) I should have brought extra silicon fittings with me (I’m not a boy, but the boy scout motto is Be Prepared after all). But I think this also says a lot about ICTD and technology transfer in general. What happens when we take information technologies and deposit them in developing regions without a thought towards their repair and maintenance?

Eventually things break, parts get lost, and things must be repaired. And spare parts don’t exist, or even if they do exist, or can be acquired, they are difficult or too expensive to acquire. So some people improvise. Others just allow the donated equipment to lie fallow until a new donation arrives, or another benefactor comes to repair the equipment. I have seen rooms full of dead computer equipment, “computer graveyards,” in Mexico, India, Ghana, and Uganda. I like mobile phone projects because there is a rapidly growing infrastructure and ecosystem outside of our own projects to support the repair and replacement of the mobile phones we are deploying – yet the truth is that often we are using very specialized phones. So even as mobile phone project developers we must be aware of the maintainability and serviceability of our devices. At least these objects are generally familiar to mobile phone service reps! (Maybe less so the android phone and the Palm phone. The blackberry seems to have gained popularity though. Obama has given it a popularity boost out here too!) Laptops are serviceable because they can be carried to a service center. But desktops are the hardest to carry because people are afraid to move them – afraid to break them, especially with the associated heavy CRT monitors.

And so while I was able to repair my headset, I have to admit that my solution isn’t perfect – they certainly aren’t the same as they were before. Getting the technology out here is one thing – making it work over a long period of time is yet another.

Tweet This Post Facebook

My Stopgap Ergonomics Setup

Well, you can’t really call it Ugandan style if it involves a stack of 4 eeePC boxes, admittedly, and I’m not actually going to use this as my permanent setup. But I thought it would be entertaining to share this with you.  Most days I’m too lazy to set up the entire stack – I only do it when I’m doing a LOT of typing, and/or my tendonitis has already flared up.

It’s basically my stack of eeePC boxes, with an open binderbox being used as a keyboard tray, and the laptop on top. The top three eeePC boxes are raising the laptop to eye level and counterbalancing the keyboard so that it is stable while I am typing.  There’s a fourth underneath (not visible) that lifts the keyboard a little off the chair so that the keyboard is a little below elbow level and such that I can sort of fit my knees under it. I’m using my goldtouch keyboard and a small portable usb mouse, connected to a 4 port usb 2.0 hub.

I’ve wandered all around looking for actual keyboard trays, and everyone looks at me a little blankly. Hah. But I finally found a store that actually sells computer desks for $70-$100 with built-in keyboard slider trays and adjustable “secretary chairs” for about $100. Expensive, especially by Ugandan standards, so I’ll have to think about it pretty seriously, but ergonomics and RSI is serious, yeah?

My other theory is to hire a carpenter to build a wooden keyboard tray into the desk pictured on the right of the picture.  Or to see if I can find a small waist-high table to fit under the desk.  I think I’ve been too spoiled by keyboard trays that can be tilted in every-which direction and fully adjustable.  Strange to have to find a way to cope out here…

Tweet This Post Facebook

So I mentioned before that I purchased a bunch of eeePCs to test out in the health clinics and to use in the management agencies as asyncronous web servers and health information management devices.

I suppose to some extent that in retrospect these clinics will have wanted Windows on these laptops so I’ll eventually have to port all of the software to windows, but for now my systems are running on Ubuntu.  It’s just easier that way. 

Installing Ubuntu was remarkably easy. There’s instructions online here, as well as lots of hints and fixes.  This is sort of my simplified version for the particular eeePCs I was working with.

Ingredients:
    1 latest distribution of Ubuntu (currently 8.10)
    1 external usb cd or dvd-rom drive (e.g. the LG-GSA-E50L 8x USB DVD-RW)
    1 eeePC (i.e. the eeePC 1000HA, 10in, 160GB, 1GB RAM, 6-cell Battery)
    1 wired Internet connection (and presumably an ethernet cable)
    Note: In theory you can also install from a USB Stick

Instructions:
    1. When turning on the eeePC, press f2
    2. Verify that a) wifi is turned on (this is for later) and b) the usb device is listed first in the boot order
    3. If the dvd device is not connected, with cd burned with the latest copy of ubuntu on it inside already, do that now, and then continue booting
    4. Install and continue, following normal instructions
    5. After install completes, update all packages via a wired Internet connection
    6. From fixes page, you will note that wireless does not yet work. You’ll want to follow the instructions there, but do not do the modprobe ath5k. However, you do need to install the backport modules:

  sudo apt-get update
  sudo apt-get install linux-backports-modules-intrepid-generic

Wireless should work after this. I used Ubuntu Ibex 8.10

I neglected to put in all my arguments for netbooks vs the mac minis we used in the Ghana consultation network, vs locally purchased desktops, vs actual servers. Basically it boils down to the fact that the netbooks have built in batteries, so we don’t have to purchase UPSes to use as backup power for then the power goes out. And we don’t have to track down a separate monitor, keyboard, and mouse every time we want to do something, which was frankly a pain, when we were working with the minis in Ghana, and couldn’t access them via the network. Laptops are designed to be disconnected from power on a regular basis, and have built in peripherals. Convenient. And actually cheaper than the minis, if a little underpowered, comparatively.

Powered by Qumana

Tweet This Post Facebook

I’ve been using tri-band antennas from Wilson Antennas, but unfortunately they don’t actually work in the places where you need the antennas because, well they’re made for the US-based frequencies (e.g. 1900, rather than 1800)

Jeff Wishnie from Inveneo points out this antenna for boosting signal strength:

This high gain, wide band, directional antenna covers the GSM900 and GSM1800 / UMTS bands. The kit contains the antenna with 7 m cable and the Universal Cellphone Adapter packaged into a sturdy box with detail instructions on installation and use. This antenna covers the 900 and 1800 MHz band which is used in built up areas.

Features:

* Broadband
* Covers various international cellular bands.
* Robust and weatherproof.

This antenna can be bought with the applicable cables for the different cards. Versions are available for:

* Vodacom OPTION Card
* Vodacom NOVATEL Card
* Vodacom HUAWEI Card
* MTN Sierra Wireless Card
* MTN HUAWEI Card
* Cell C NOVATEL Card

Powered by Qumana

Tweet This Post Facebook

1 solar flexible-panel charger (retail $99-ish) for 4 AA batteries
2 solio solar chargers, for mobile phones and usb devices
(can be charged via solar or grid power. I only have one adapter
for grid power) (retail about $80 each)
1 inverter (retail about $40)
1 potenco pull cord charger for charging car batteries (on loan)

plus lots of portable cell phone chargers (that can double as flashlights, $25ish) using AA batteries, and a few $5 LED book lights with hard-to-replace batteries. I also have two sidewinders (retail $25) which are wind-up flashlights+cellphone chargers. They aren’t
particularly useful for charging cell phones, but they are better than nothing in a pinch.

the potenco pull cord charger is also a lot of work to charge a car battery – you wouldn’t want to use it unless you could hire some kids to pull at it on a regular basis for cheaper than you could pay for the equivalent work’s worth of fuel. It’s a great device tho, and
could easily be hooked up to a bike or a windmill (not too fast of a windmill tho). To charge a car battery, basically you bolt the charger to a wall, grab the two ends of a rope, and then you swing your arms like you are on a nordic track. Some lights come on and the
battery starts charging.

I haven’t tried it yet because I don’t have a car battery yet. I’ll try it next week when I go to Rubindi and I’m staying in the rural health clinic overnight. By some miracle I haven’t experienced a single power cut yet.

It’s also been raining, so the solio chargers and the AA batterycharger don’t seem to have charged fully yet, although maybe they would if I put them outside directly instead of just in my window..  I suspect the AA battery charger’s 4 hour claim for 2 batteries only works for the standard AA rechargeables and not my super-high capacity 2700 mAH batteries, because they only every charged up halfway.   In the end they really just work well as capacitors for grid power. =P

Powered by Qumana

Tweet This Post Facebook

It’s part of an interesting new direction for technology research – instead of just building faster, more high-resolution (and more expensive) devices, people are working on ways to build low cost devices that are more robust, can be mass produced, and can provide good enough information for primary triage.

On another note, these devices (as the economist article posits) could be well deployed with a good mobile-phone-based data collection system – collecting not just text and numbers, but images as well.

As part of the evaluation for the Uganda OBA project, Ben Bellows and his collaborators at Makarere University are conducting a household survey in the coverage area of the project and in a similar control area. As part of this survey they have to also do sexually transmitted infection (STI) testing, trying to determine the actual prevalence of STIs and not just an estimate based on who comes in for diagnosis and treatment. Can you imagine how much easier and verifiable these surveys would be if 1) the data collection could be done electronically, and 2) digital media for the testing could be integrated into the data collection records? Not that all diagnoses could be done with cell-phone microscopy, and you still need careful sample and slide preparation. But it’s still something to think about…

Tweet This Post Facebook

Tweet This Post Facebook

Of course – before everyone goes off running to deliver this very fine product to the masses of doctors in developing countries, there are a couple of catches. The data is very much geared towards doctors in the US, which has a number of implications. 1) Only the drugs that are available in the US are listed, and then with the US names. So, for example, many of the newer malaria medications which haven’t been approved in the US (like Coartem) won’t be there. And other drugs like paracetamol (as it is known in the UK and former British colonies like Ghana and Uganda) will be listed as acetaminophen. 2) There’s a yearly recurring cost of $100. Of course, this might not be out of reach for these doctors, and certainly is about equivalent in price to the paper versions of these reference guides, and about 1000 times more portable.. 3) The pathology is different – a doctor in the US wouldn’t expect TB, where a doctor in Uganda or Ghana would know to look for TB symptoms.

I bring all this up because I spent part of this past summer in Uganda with the 2007 East Africa Blum Fellows visiting some of the Uganda Health Information Network (UHIN) deployment sites in Lyantonde and Rakai. What struck me most was not the specific programs offered by the project (digital submission of health outpatient statistics, and dissemination of malaria and pediatric health information), but rather how they appropriated the devices, installing and sharing their own applications, and using the Excel application to track inventory and patient logs. They just drink up this data, reading whatever they can get to learn more about how they can care for the wide variety of conditions they see every day. So.. in addition to whatever information management functions I can put into place, I hope I can also help put more information in the hands of the doctors and clinicians and nurses I’m working with.

There’s of course still a lot of other issues to deal with – everything from power for recharging to the cost of the devices themselves (~$70 for a Palm, and $300 for a Palm+Mobile Treo) to maintenance and sustainability. I still want to try putting this type of information in their hands, with all of the appropriate warnings, as well as more locally specific information, like local health bulletins or Hesperian’s translated Where There is No Doctor series. Let me know if you have any suggestions for mobile health applications!

Tweet This Post Facebook