We are building a company (<a href="http://www.keedo.tn" rel="nofollow">http://www.keedo.tn</a>) to reduce the gap in education accessibility between urban and rural zones and private/public systems. We provide web apps and learning tools (we already started developing some tools like a virtual piano, body anatomy viewer, maps for geographic subjects, stories reader and so on) and we are fully cloud based, so everything is accessible and shared over the network.<p>Our big challenge is to provide a cheap computer with internet connection, (wifi/3G) support and acceptable performance. For information, the family income for our first clients in such regions, is on average $130 per month.
We love the OLPC_XO <a href="http://laptop.org/en/laptop/" rel="nofollow">http://laptop.org/en/laptop/</a> project idea (worldreader.org also) but their price is still high for our target clients and we try to avoid donations as much as possible.<p>First we thought about low cost white label tablets (most are china ODM/OEM providers), but from experience and reviews, they seems to be less performing with intensive usage.
Now we have two solutions, either we use an already proven SOCs like PINE A64, Raspberry pi or we develop our custom pc.
Any pointers, suggestions or collaborations will be highly appreciated!
I would encourage you to forget about hardware and concentrate on the software and applications.<p>Based on my experiences (I work for an ISP but we have a "sister" company that's an MSP that provides IT services to, among others, several K-12 schools in such rural areas), most of these schools either 1) already have available hardware (iPads, Chromebooks) due to existing 1:1 programs or 2) have decided to not (yet) pursue that (usually because of financial reasons).<p>In the former case, they already <i>have</i> hardware and don't want to buy even more. In the latter, they simply <i>don't have the money for it</i> (it'll be an even harder sell if it's not "general purpose" and can only be used with your stuff). Even if you manage to get it under $50 per student, that's still a lot of money that they could likely spend elsewhere.<p>Develop your applications and concentrate on targeting the most popular platforms (iOS and Android) and forget about trying to sell them hardware they can't (easily) reuse. Definitely forget about cellular connectivity and stick to Wi-Fi instead.
So, from the .tn URL and family income listed, I am inferring this is based in Tunisia.<p>Can you optimize this for a cheap smartphone? I think that would probably be a better solution.<p>I am American. I was homeless for 5.7 years and I began learning to make money online while homeless. This is part of how I got back into housing. I have some experience with various tablets, laptops and phones. Phones have proven to be a surprisingly robust internet solution for me.<p>I currently have a $30 smartphone (brand ZTE). I have Tracfone service with it and I can use it on Wi-Fi. Although the screen is tiny and storage is limited, it has overall better performance than the $70 tablet I have. I rarely get on the tablet. My sons use it, but I spend almost all my time on my phone at home, plus I go to the local library to access a proper computer.<p>From what I have read, phones are a much bigger thing in Africa than tablets or laptops. I have read that everyone in Malawi has their own sim card even though an entire village might have a single phone.<p>I can do a surprising amount on my phone. I can blog, take photos, play games and do online banking. I have to be mindful of storage limits and I have to periodically clear my cache and do storage management and I have to be a little pickier about limiting my apps to the essentials. But it does a surprisingly good job of letting me do most of the things I want to do online.<p>When I was homeless, battery life was a big deal. Large screens on tablets really burned up the battery life. A phone or small tablet was generally better in that regard. Devices with good battery life were vastly more useful. I had access to electricity during the day most days, but I was on battery power at night. How much I could do was very dependent on battery life.<p>Large parts of Africa will have limited access to electricity. I have read that in some areas, one person will collect up all the phones in the village once a week and hike to the nearest paid recharging station and get them all charged.<p>So I will suggest that a cheap smartphone with a long battery life is probably the best solution for this project. Then design your program with those constraints in mind.<p>Edit: I will add that a $200 laptop was generally worse than a $50 tablet. We had to spend $300 on a laptop to get one we didn't hate.
My thought? E-waste.<p>This is how I got my first computers; obsolete and broken computers cast off by people wealthier than I was. [1] It's great, because you get to learn how to fix the computer as well as whatever it is you learn on the computer. I imagine my life experience is super different from those you are helping, but I can tell you that early experience fixing obsolete computers has vastly helped my career.<p>It <i>is</i> a lot more labor intensive, because you have to have someone to teach the kids how to fix the damn things, (for me, my dad and a copy of minasi's "upgrading and repairing PCs") but once you have that skillset locally? it's probably sustainable, because the rich will always be throwing out last years gadgets.<p>In some ways it will be way easier for you than for me; when I grew up in the '80s and '90s, there wasn't a lot of standardization, so with my 'catch as catch can' hardware acquisition strategy, I'd have to completely change my software stack every time I got new hardware. These days? Most educational software runs in a browser (I use and heartily endorse Khan Academy for the parts of learning that can't be done from a book alone) and so you can have a diverse fleet of hardware and even operating systems, and as long as you have enough local skill to bring the things up to the point where they can run a browser, you should be good to go.<p>The big problem with this plan is power; Nicer gear from the aughts ought to run a browser just fine, but it will chew up a lot more power. If you have to pay for unsubsidized electricity, more modern gear might make more sense.<p>[1]There was also this really juicy (for me) sense of acquiring "means of production" - I <i>owned</i> these computers that others had cast off as worthless or broken, and I turned them into a useful resource for myself. I think that if you could work something out so that the kid in question gets to keep the computer they fix... that might be extra gratifying. I know it was for me.
As you likely know, you will need to be <i>very</i> careful when developing web-based applications for the resource-constrained devices you mention. A typical bloated web app as made by most companies will often not run well in resource-constrained environments like many ARM SOCs. I would encourage you to approach the app design process with an extremely minimalist approach, striving to avoid libraries and use vanilla JS as much as possible, use tiny assets, and keep allocations to a minimum (maybe consider object pools for certain applications). If you use a minimal framework, Mithril JS might be a good choice, or domvm, or something similarly minimal. You probably want to stay away from React and Angular.
If you're looking at any kind of volume, you should be able to get Chromebooks for under $100 direct from China. When I was at a K12 school about 3 years ago, our purchase of ~250 Chromebooks allowed us to knock nearly 25% off the per unit cost to get them down to ~$150/device. And that was going through a third-party that was white labeling the devices.<p>The machines we ended up with were 2GB of RAM/16GB storage devices with 11" screens. Basically they were the reference designs Google released a while back when they opened up Chromebook manufacturing. And for the 4th-12th graders we gave them to, they were amazing.
Recovering olpc volunteer here. Happy to hear of your labors!<p>One of the concepts a related project explored was using live / persistent USB disks to preserve student ownership and facilitate unscheduled explorations.<p>Multiplexes hardware across students at different times, facilitating students using a consistent environment at home on old P4 desktops and at school on whatever's available.<p>Even large (32+GB) portable storage devices are available under $50.<p>Check out "sugar on a stick" and "open1to1" for related trains of thought.
We are working in Zambia to building tools for people who want to create their own electricity supplies using local resources.<p><a href="https://localelectricity.org/" rel="nofollow">https://localelectricity.org/</a><p>We are also working closely with a local NGO who do coding workshops for rural communities.<p>Would love to chat and explore possible collaboration opportunities. Email is Samson [at] localelectricity [dot] org
I don't know where you are thinking about implementing your"solution". There are a lot of people trying to solve the problem you are solving, even governments have. I am writing from Ghana and the government has tried two times and failed. Two questions you need to address: 1. does your solution really align with the people's need ? 2. Is it sustainable? After you have found interest, you ask if the people are also interested in managing the program, after you are gone. Is there the required infrastructure to support. I am in development work and this is really core to making impactful change, I was going to say lasting change but I took it back. There are other level of questions but please answer the two questions above before you even move ahead with your idea otherwise it's a waste of resources.
I think this is amazing, and credit should be given where credit is due. Are you basing your concept on learning management systems like moodle. Is that what you are? I know, you're probably so much more. But I need to ask, how do you organize content that's already available into bite-sized lessons easily digestible by children? And it's easy to get muddled into providing content, along with a medium for that content. And then you go into grade levels, standardizing, not to mention having proper internet access, along with power. I mean how do you do it?
You should reach out to the folks at learningequality.org. They focus on totally offline use cases but have a huge global partnership network and could definitely give you hardware advice.
Meta question, do all text submissions get to include active links in the text or is that handled manually by the HN team? I noticed some Launch HN: submissions over the past year have been doing it but AFAIK this is the first for a non-YC company.
Refurbished Chromebooks would be hard to beat. Lots of sub $200 choices, and has the keyboard and monitor built in. You can also download a new Seabios ROM for most of them and run regular Linux.<p>Affordable rural internet access might be a bigger hurdle.
Nice to see something new going on in EdTech space! I'm the CTO of bibliotech.education, a subscription textbook company and happy to have a chat regarding collaboration or contacts in either the US or Europe.
Design the app to work in smart phone. Families might benefit have cheap smart phone for multi purposes. Cheap smart phones will help students as they grow up to familiarize the mobile technology.
I know of a company Learning Equality(<a href="https://learningequality.org/" rel="nofollow">https://learningequality.org/</a>) they also work in this space.