I am Mike Dawson of Paiwastoon and I’ve spent 6 years working with Education Technology in Afghanistan. There is of course no shortage of challenges with the country, yet in spite of them, there is progress. We now have 50%+ mobile penetration and millions of girls and boys back in school.
But I still face roadblocks, literal and figurative, every day. I can’t get an experienced team of software engineers in Afghanistan. Electrical power keeps on playing games – on, off, off, on. The Taliban keep shooting rockets at our KFC in Kandahar, often when I really need french fries. And with the American withdrawal in 2014 it looks like we will never see a real Starbucks coffee shop in Kabul.
So when it comes time to deploy a ICT4 Education project, I really need something that just comes out of the box with instructions, and I can get anyone to assemble, Just like an IKEA coffee table. Preferably with no little screws missing.
Many in my target audience already have computers with 2GB+ of storage, 200Mhz+ processors, 32MB of RAM and small screens that cost around $30. These marvels are called feature phones. I have learning content that I want them to use: questions, games, videos, audio. But where is the toolkit to connect the two?
If I wanted to do an SMS based project of some kind, I’d just use a cloud offering or download FrontLineSMS. If I wanted to do mobile data collection there’s almost as much selection there as there are types of cereal available in the supermarket.
But what platform do download when I want to use all these mobile computers (phones) as computers for learning? Why is it that putting together the computers people already have and quality learning content is more complicated than browsing and selecting from the IKEA catalog? Even with impossible to pronounce Swedish names?
I find this simply unacceptable. We need a toolkit and templates for mEducation, we need to stop re-inventing the wheel every time we do a pilot. We have mEducation pilotitis, and re-usable toolkits are the treatment that allow us to rapidly fix them, change them, and scale them once they work. Otherwise we just go back to the drawing board for months of planning and re-engineering.
Building the tools
So we are building Ustad Mobile (“ustad” means “mobile teacher” in Dari/Pashto) which will enable anyone who can use word processor or powerpoint software to make interactive courseware about any topic they want, then make a smartphone-like experience without needing connectivity, using any feature phone or better that happens to be available.
Finally we will be able to deliver radically enhanced learning experience without giving out expensive new equipment.
We’re getting there – we’re not there yet.
Initially we coded. We added more interactive learning templates to it (hangman game, sort exercise, place/find objects according to a hint). At this point we had one developer, who is also the one writing this article. Supposedly he also had a part time job as CEO of the company. Our office was a virtual factory assembly line of powerpoint slides, teacher designed wireframes and sound recording.
Using EXE Learning our teachers and content adaptors (who do graphics, sound recording, etc) put together a whole version of the Afghan literacy curriculum for the illiterate, then put it on SD cards so that the app runs on the feature phone.
We are live field testing
We recently provided the app on Nokia phones to women in an existing Ministry of Education class to use for homework. We tweaked some of the interactions to make them easier, we started stabilizing the authoring environment and the app itself.
Then we got a new test group: Illiterate policewomen sponsored by the Dutch Embassy overseen by the UN Assistance Mission to Afghanistan.
We tried improving our statistics gathering. The idea was to bluetooth data from students phone to teachers phone. This required users to push “OK” within 30 seconds of activation. Fail! We switched to sending pre-compressed logs over GPRS. Win!
This M&E process works even when connectivity is unreliable, uses around $0.15 of credit per month per student monitored, and every 5 mins if there is a connection we get the data automatically in the cloud.
Now I’m seeing that the policewomen are clocking up combined total usage of 10hrs to 30hrs a day across around 100 devices as they practice literacy. The phones send logs every 5mins when they are in use.
What’s still to come
We’ve tested it on Nokia phones, but “should work on anything” and “actually works on anything” are two distinctly different concepts. So now we’re doing compatibility testing and engineering supported by Khalilzad Associates so it’ll work on feature phones, and via HTML5 for Android, Blackberry, Windows Phone and iOS for all those impoverished masses who are rushing to buy the iPhone 5C.
While cross platform capacity will be released mid to late November 2013, I haven’t found a single instance where using more advanced hardware really worked in education.
We shall work tirelessly to make the world of mobile learning more like the world of great value flatpack furniture. Implementers who know their audience including government, NGOs and the private sector should be able to use the potential of the technology at their disposal without needing expert carpenters and consultant interior designers at every step.
When the toolkit is in the field, for mobile learning, and hopefully soon in other areas around democracy, anti-corruption and more, impact can be achieved at scale as implementers need not re-invent the wheel. They only will need to read IKEA instruction manuals.