Last year we (Gertie Clabbers and Harrie Vollaard) travelled for 6 months through Asia and visited several OLPC deployments in Mongolia, Nepal and India and started one in Russia. It has been a trip for a good cause.
Gertie and Harrie donating an XO
With our charity foundation Making Miles for Millennium we raised funds in The Netherlands to support the different schools with XO's or infrastructure.
The preparations of the trip started however already long before our travel. Because we decided to specifically help OLPC deployments, we also have a relatively long experience with OLPC as an organization. We have written down all our experiences of working with OLPC in: Evaluation of and recommendations for OLPC 2008
Report Goals
Our main goal of the report Evaluation of and recommendations for OLPC 2008 is to recommend OLPC how to improve their organization. Especially now with the recent cuttings it is even more important. Our belief is without any change of OLPC, OLPC won't last for long. If they don't change the way they are organized, they will never be a big player and loose the commitment of their supporters. The XO is only a device that sparked up the process in the developing countries, but is not the device that will last. Competitors will come and will do a better job in delivering and support.
We shared the report with OLPC. They asked me to publish it on OLPC News because:
"We would like to be able to respond to the points and questions you've raised in your evaluation in a public forum. We think your evaluation is very good, and well-placed, and that you and the rest of the grassroots community deserves transparency from OLPC on how we go about answering those questions."
We also think it is a good idea to publish the report. We hope it starts up a discussion about the recommendations and finally gives:
- An idea/impression if the community agrees or not with the issues and recommendations.
- Improves the recommendations because of comments of the community.
- Creates openness and transparency from OLPC how to tackle issues and to improve their organization.
Report Conclusion Synopsis
The XO is great, but the organization OLPC can be improved. The organization OLPC tries to switch from a pure research organization to a supplier of the XO when they first started to deploy the XO in 2007. However after 1 1/2 years it looks like OLPC still has no business processes in place. The people who work with OLPC have no experience with these business processes and do not know how to organize a nonprofit organization into a streamlined organization that can handle simple orders. After all it is only one product OLPC 'sells'.
In most countries we have visited OLPC's role is a pure supplier's role. The way they act is however differently. Instead of making it very easy to order XO's in larger quantities, you have to discuss why etc. In addition, what doesn't help either is, OLPC hardly answers any email. The way OLPC is organized and how it communicates, it still looks like a research organization.
Our Recommendations
- Have a clear focus as an organization. Determine if OLPC supports only developing countries or everyone.
- Support small‐scale projects and small numbers.
- Calculate the same price no matter the quantity and lower the price.
- The core competence of OLPC is research & development. As done in the consumer market with Amazon, look at a global company that can handle and deliver globally for the GiveMany program. The advantage for OLPC is, it can still be a research organization. There is no further need to streamline the organization and implement business processes.
- Let local organizations, NGO's determine where XO's can be deployed.
- Work together with NGO's on local level instead of building up how to deploy a project in a developing country by yourself.
- Have the authority to decide as low as possible in the organization.
- Solve the technical support and guarantee issues.
- Embrace the grassroots and supporters.
- Communicate and answer your emails.
- Gather and organize all the documentation needed for the deployment on the Wiki.
- Work with a clear functional roadmap and publish it on your website.
good-intentioned, but very un-realistic set of suggestions.
Yes, they are all *apparently* sensible suggestions, but they can't be put into practice because, much like many others before, ignore a simple fact: the XO is NOT a viable product.
It is not viable (as in "un-sellable") because it is incomplete: it lacks most of the features that promised to make it stand out from the crowd. two years later, there is no power crank, battery is not "measured in days, not hours", source code is not available to users, the mesh network is unreliable, Sugar is a mess and the price is twice (soon to be quadruple!) the initial tag.
I say it clearly: before you think you can compete, you must create and FINISH a good product. No amount of enthusiasm and good intentions will ever override reality.
Yes, I know...it's harsh, but it is also true.
I agree with you on that features can be improved. We mentioned e.g. the accessibility. The XO is however not that bad in the field is our experience, esp. in comparison with other devices (no moving parts, robust device etc). By the way I'm talking about the situation in developing countries.
The evaluation is however focused on OLPC as an organization. More curious about thoughts on these issues/recommendations.
I actually think the suggestions are spot-on, and largely achievable, but it will take a reduction in the ego-driven nature of OLPC to step back from it's prescription of who can do what with the XOs and to focus on making the XO technology top-notch (and - gasp - competitive with the 4PC market they've created).
It has a myriad of features that make it stand out, even if the crank didn't work out (I *STILL* get questions about the crank), and the yo-yo remains "not available for purchase."
This is part of why I continue to push for a commercially released version of the XO, so that organizations can purchase them and do with them what they like.
Don’t you think if OLPC is completely commercialized or bought by a third party, they will lose the connection with the grassroots; their unique selling point?
Another possibility would be to refocus on research and development. 30 people would be more than enough (Invest in developers instead of consultants.) Look for a commercial party/resellers or whatever for the distribution of the XO. At least you get rid off the single point of contact. And more important leave the deployment to local NGO’s.
It's not that bad. While the XO is not what was described in January, 2007, it is a viable computer. Its battery life is measured in hours, not minutes, while doing serious computation-intensive activity; its mesh networking works well enough for small-scale groups and its ability to connect to wifi is not terrible (though imperfect).
The software needs work. (What software doesn't?) As an R&D outfit, OLPC in combination with SugarLabs *is* working to improve this.
IMO, what's lacking is attention to enabling non-OLPC/SugarLabs developers. I have found the process of creating software for the XO to be extremely difficult due to utter lack of attention to questions posed on the developer mailing list and on various foura. (See point #10.)
I am not a clueless newbie when it comes to software development, but while working with the XO, I sure feel like it! Point #11 hits home for me.
1) Check to see if Apple has an interest in taking over the project or contributing in some significant way.
2) Plead to Warren Buffet for funding.
Point #1 is the crucial one and that's where recommendations should be targeted. Everything else follows.
OLPC looks like (is) an academic experiment (thus the tightly controlled conditions and environment) dressed as a social cause/movement/goal.
It just needs to address the impact of specific predefined variables in a hypothesis made long ago and still missing scientific proof. Everything else is just a very important, but just the same, "dressing".
The question that is actually asked is "does a large deployment of 1:1 computer assisted education improve learning and social outcomes in underdeveloped countries?"
I guess if they can keep it up an a couple of countries will have their answer.
What they need to be convinced of, is that with the current course and mentality the experiment will never come into fruition and most likely abandoned too-soon-to-be-conclusive by the early adopters.
It is a valid and very interesting experiment but it can be much more than that.
The sooner they decide on their priorities (either way) the better is going to be for everyone involved.
I hope the answer to point #1--"Determine if OLPC supports only developing countries or everyone."--is "everyone". It's silly not to take advantage of rich English-speaking countries whose people will buy the laptops and enlarge the community working on the software. Even though the end goal may be to help developing nations, it makes more sense to harness rich nations for their money and free time than to expect poor nations to supply all the money and time that the project needs.
And, with regard to the issue mavrothal brought up, the effectiveness of 1:1 computer-assisted education can be studied just as well in a rich country as a poor country.