In Work Years, Recycling Beats New OLPC XO Production

   
   
   
   
   

Are you are still wondering how Nicholas Negroponte got to a One Tennis Court Per Child analogy for traditional computer labs?

While you (and I) are still befuddled, did you pick up the complete dismissive tone of Nicholas Negroponte when discussing recycled computers? That is older standard desktops or laptops sent to the developing world for use in computer labs there.

Not only do computer labs not pass his "pencil test" where he says that:

One does not think of community pencils-kids have their own. They are tools to think with, sufficiently inexpensive to be used for work and play, drawing, writing, and mathematics.
There is also a new rationale specifically against recycled computers on the improved OLPC website:
Finally, regarding recycled machines: if we estimate 100 million available used desktops, and each one requires only one hour of human attention to refurbish, reload, and handle, that is forty-five thousand work years. Thus, while we definitely encourage the recycling of used computers, it is not the solution for One Laptop per Child.
Whoa, of all the arguments One Laptop Per Child could use to bat down recycled computers, they are going to run with "work years"? No wonder they need W2 Group marketing savvy. The "work years" angle is crap.

For argument's sake, let's agree with OLPC and say a recycled computer "requires only one hour of human attention to refurbish, reload, and handle." An OLPC XO requires at least that, if not waaaay more to build from scratch. How many man hours does it take to make each component of the OLPC? And then assembly? Quanta is using an entire factory to assemble Children's Machine XO's. There is a whole other factory (or factories) to produce the screen.


Where most CPU donations go

If OLPC (or just Negroponte) want to argue Children's Machine XO's are better than recycled computers (which I agree they are), they might want to use the real-world reasons why many of us who implement technology in the developing world shudder when we hear of a computer recycling program. Reasons logical, practical, and unlike OLPC's, valid:

  • Recycled computers are often non-working, end-of-life junk
  • They have wildly different hard/software configurations requiring manual inventory
  • Power-hungry Pentiums don't live long in low-voltage, high power-spike environments
  • "Free" computers donations can destroy local PC assembly businesses
  • There is usually no provision for support and maintenance - the real cost with technology
  • "Second hand" can imply the recipients are "second class"
Now that's not to say all computer donations are so ill-suited. There are outstanding examples, like SchoolNet Namibia but they are painfully rare. And even rarer still is any logic to the OLPC "work years" argument against recycled computers. Or the One Tennis Court Per Child analogy for that matter.

So OLPC, do us a favor, better your logic in your arguments, especially on topics we all agree on.

Related Entries

9 Comments

Hello Wayan

An interesting article, and I agree to that "The "work years" angle is crap".

You state:
Recycled computers are often non-working, end-of-life junk.

Answer: It is no problem to sort out the computers that are in good shape, make "plug and play-packages" as thin clients and then send them with containers to Africa and elsewhere - as FAIR do. We never send junk.

You state:
They have wildly different hard/software configurations requiring manual inventory.

Answer: This is partly true, yet has not been a problem for us.

You state:
"Free" computers donations can destroy local PC assembly businesses

Answer:
This is not the case. The schools that are helped in this way would never get ICT in any other way. ICT is close to 3 times more expensive in their countries than in US and EU. On the other hand, when pupils come out with ICT-education they will stimulate the local ICT-business.

You state:
There is usually no provision for support and maintenance - the real cost with technology.

Answer: This can be solved with FAIR's model: FAIR offer a holistic approach to the transfer of technology and knowledge between the north and the south, including hardware, software, installation, training, support, maintenance, software development, take-back system for EE-waste, advisory and last but not least – financing. Financing means that FAIR offer a financial model where a combination of the government, companies and private persons in the country of origin cover all the costs for the transfer of technology and competence to the recipient country in the south. The reipients are not charged for FAIRs products and services." For more information;
http://www.fairinternational.org/fi-artikkel_244.nml

You state:
"Second hand" can imply the recipients are "second class"

Answer:
This has been the result from several players which work with ICT-aid. But when 3 years old pentium 2-3 are used as thin-clients in a proper way, it is First Class for 4-7 more years, linked to a server which is good enough. The recipients will be/are ensured that this way is the best way because of good references in the same, or other countries.


best regard
Terje
www.fairinternational.org

Terje,

I'm glad that you spend so much time and energy on your recycled computer model. I only wish that were the norm.

The norm is a shipping crate full of unsorted, untested, often unworking hardware that becomes a technology albatross around the neck of the non-techie recipient organization that expected easy plug & play, fully supported computers.

Hello again Wayan

You are right about the fact that IT-junk is shipped as aid. That's why FAIR has established our high standard, and urge the world community to adapt.

If EU and US adapt our model, the result would be that all the poorest countries in the world could have an up to date ICT-standard in their schools - for free !

Wouldn't that be beautiful?

And if OLPC are used in “One PC-lab per Per School”, it would also be a big help for the poorest countries to get there much faster.

Best regards
Terje

A couple of years ago Salon had an interesting article about the Alameda County Computer Resource Center (http://www.accrc.org/):

"Building the underground computer railroad" (http://www.salon.com/tech/feature/2002/09/23/antiglobal_geeks)
"Anti-globalization activists in Oakland, Calif., are recycling old machines, loading them with free software and shipping them off to Ecuador."

I may be wrong but I think the above doesn't require viewing an ad.

Unless the people (and robots!) at the Quanta assembly line have the skills needed to recycle old computers, the OLPC observation is indeed valid.

Recycling is a worthy effort (I have done it myself a few times), but it just doesn't scale.

Scale is exactly the word. Hand-refurbishing old PCs is not directly comparable in any meaningful sense with mechanized production out of a tooled-up factory. "Economies of scale" was/is the whole point of the industrial revolution, right?

To Jecel Assumpção Jr and John Maxwell

"but it just doesn't scale."

In what way do you mean ?

Logistic, funding ...


"One PC-lab per School" (OPCS) with reused ICT-equipment can scale easily in every means, for the best of the environment and the digital divide.

As you now, it is in an environmental perspective far better to continue to use for 4 to 7 years, than terminate 3 years old very well functional ICT-equipment

Because FAIR reuse ICT as thin-clients with Linux, companies with a FairRecycling-certificate can document the far highest reuse-percentage in the market. They can also document that they make a difference for the school- and health system in the third world. This is good Cooperate Social Responsibility (CSR) and look off course also good in public.

Good measurable CSR-programs are companies increasingly willing to pay for and this is financing the part FAIR needs beside official funding and other sources. (More info in our Annual Report in the upper right corner at www.fairinternational.org. )

Among many others has WWF-Norway, Amnesty-Norway, The official “pollution monitor” agency (www.sft.no), “the governmental consumeragency”, one university, 2 city's, 2 hospitals – big and small companies in most branches has chosen the worlds best way to give further life to their ICT-waste. (complete list at website)

150 - 200.000 thin clients for the third world is the full potential from Norway, and when Europe and US joins up will Africa, South-America and part of the Far East be covered in the school system included parts of the universities and health institutions.

If Negroponte and OLPC simultaneously worked with the same, and use the OLPC as thin clients in OPCS, the goal would be reached much faster.


From our press releases: ( You can find them at www.fairinternational.) FAIR offer a holistic approach to the transfer of technology and knowledge between the north and the south, including hardware, software, installation, training, support, maintenance, software development, take-back system for EE-waste, advisory and last but not least – financing. Financing means that FAIR offer a financial model where a combination of the government, companies and private persons in the country of origin cover all the costs for the transfer of technology and competence to the recipient country in the south. The recipients are not charged for FAIRs products and services.

There are no part of our way to handel ICT-aid that we can not scale to any propotion.

And the solution is FREE OPSC for the poor with the technology which is compatible with ours. The sadest part with OLPC that many seems to forget, is that even when it seems cheap, it is an ocean of money for the countries that is the target group. And as you know, they also have other fundamental needs that must be solved also.

With the development countries in focus


Sincerely yours
Terje Tjensvoll
FairInternational.org

Terje, by "doesn't scale" I mean that something that works perfectly for 50,000 machines might break down when you try to deploy 500,000. For example: as long as you (and everyone else involved in recycling) are working below a certain level there will be an endless supply of used computers that will just keep getting better and better each year. But if you go above that level then you will be forced to expand your search to ever older machines and these will have lots of problems.

The newer machines will have an Ethernet port compatible with the PXE standard, for example, so using them as a thin client booting off the net will be trivial. For older machines each case will be different and some will require you to find an odd file somewhere and program an EPROM chip with it to stick into a rusty old card.

And you will need ten times the people to handle all this and no matter how good your program is, the new ones simply won't have the experience and training of your core group. But as I said above they will be facing a much more complicated problem.

As for the scale of our discussions, it is "one computer per child". You may not agree with that, of course, but "one lab per school" is an entirely different project.

About using the XO machines as thin clients in your project - why not? You don't need Negroponte's collaboration for this and not even his permission. The XO can already do this as it is today (but you have to drop into the terminal screen and type in some commands) and you could easily package it up as a nice ThinClient activity for Sugar. I don't see the two projects as mutually exclusive at all.

Jecel Assumpção Jr

Our experience from using old computers as thin clients, is that we can use quite old machines, and we believe that “the certain level” that you refer to is not an issue. As you point out, new machines are getting better and better each year, also as recycling-machines, which now a big problem as EE-waste. Let us reuse this potential from EU and the US, for the sake of the third world, and the environment, that means at least 5 mill excellent machines every year.

To scale the need of manpower is also easy, and we have no boot problem with older machines today, and can not see any reason that this will give us any hardship in the years to come.

When you say that “2one lab per school" is an entirely different project.””, it is right and wrong. It must also bee about priorities when you address countries with the biggest cash problems. I believe that poor countries which have the goal to get out of the digital pit hole, should follow the example of the western world, and focus on giving the students on the levels before university level ICT-skills first. OLPC will drain this countries on recourses they really do no have, by selling them a high tech learning gadget to small children, which, when moving into next step of education, not will find any high tech or ICT at all. Therefore we gave Mr. Negroponte the idea to rethink the olpc-strategy, and focus the XO as thin-clients in “One Lab per School” instead, and have the focus on the real need this countries have. To get out of the digital abyss, they need to give ICT-skills to the young between 12 and 18 as soon as possible, and not children between 8 and 12. The risk is great that OLPC will delay to solve the primary need of this countries, in their education system, and in all the other sectors where they also are way behind the rest of the world.

OLPC or our way to recycling ICT, is not in a direct conflict, but I believe that OLPC work against the UN and worlds goal to get the third world out of digital divide. The reasons is as mentioned, because of draining the poor of funds and misleading them out of focus in how to get there much faster.

Your technology is superb, it must only be channeled differently – is my opinion.

Best regards
Terje Tjensvoll

Close