One Laptop Per Child – a preview, and a request for help

I took a day off from this year’s Pop!Tech conference to hang out with some friends in Portland. But before driving from Camden to Portland, I dropped into the Opera House to check email and bumped into Nicholas Negroponte, who’d given a talk the day before on his work to produce a laptop that costs less than a hundred dollars.

Negroponte was an advisor to Geekcorps and was extremely helpful to me as we figured out whether the organization would be supported by corporate sponsorship, foundations or government largesse. So he knows about my long-standing interest in technology in the developing world. He asked whether I was interested in coming over to the lab and seeing a demo of the machine, and talking about strategies for deployment.

Hell yeah.

The demo was yesterday afternoon, and while it didn’t include a functioning prototype, I learned a great deal more about machine than I have from previous articles, or Negroponte’s talk at Pop!Tech. He was able to answer a whole set of questions for me, and raise an entire set of new ones, which, I suspect, will take a number of years to answer accurately.

First, the name. I’d been calling the project the sub-hundred dollar laptop… the acronym of which is the unfortunate “SHiL”. Negroponte’s now calling the project OLPC – One Laptop Per Child. It does a better job of defining the project, I think – not taking the bottom out of the consumer laptop market, but providing a learning tool for students around the world.

On to the machine. While the actual prototype is being actively banged on (in preparation for a live, but tethered, demo at WSIS on November 16th), Negroponte keeps a cardboard mockup of the machine on the conference table in his office. It’s a clever little thing – I had a hard time putting it down after picking it up. You can see a design close to the prototype I saw on the front page of Design Continuum’s site – they’re evidently doing the case design for the machine… and, actually, pretty far from the design reported on in the AP story about the project.

The mockup I saw was about the size of a large paperback book. There’s a stiff rubber gasket around the edge of the machine, which can double as a stand. The keyboard on the mockup was detachable, but will probably fold out on a hinge. The system is designed to work in three modes: laptop mode (screen up, keyboard down, handle behind as a stand); book mode (screen on the front, keyboard on the back, comfortable indentation for holding it in the left hand. Pressing on the keyboard “accordian-stype” – as Negroponte puts it – allows for page scrolling); and game mode (screen in the front, keyboard in the back, held sideways, like an oversized PSP. Two trackballs, surrounded by four way buttons, on each side of the screen act as controls, and function keys on the back act as additional buttons.)

Unlike in the prototype featured in the AP story, there’s no large gap between the screen and battery section, designed as a handle. While it looked very cool, it was also a bit too fragile for the conditions being considered. The handle now is either the rubber gasket or the indentation in the back. I wonder if the hinges are going to be a problem – the current design requires a hinge for the gasket and a separate hinge that allows 340 degrees of freedom between the screen and the keyboard. Negroponte pointed out to me that the hinges in the average consumer laptop cost about $5 each, and have a complex clutch mechanism that allows a screen to be smoothly rotated and repositioned – it will be interesting to see what can be done on a machine where the hinges will need to cost pennies, not dollars.

The keyboard on the prototype I saw was removeable – I think this was a nod to the idea that separate keyboards will need to be produced for different markets. In China, the appropriate device might be a stylus and pad rather than a keyboard, making it easier to enter ideograms.

Much of the conversation about the laptop has centered on the display, one of the most expensive components of a modern laptop. Early designs called for an e-Ink display – while Negroponte keeps a large e-Ink display on his conference table (about twice the size of the one in the Sony Librié), he tells me that eInk is for the second generation of the device. (It’s easy to see why e-Ink is compelling. Not only is the display extremely crisp in black and white, it remains crisp and readable when the display is turned off, as the molecules that make up the letters retain their orientation despite the absence of current. This makes power requirements for texts that don’t change continuously very, very low.)

Instead, the first generation of devices will use an LCD screen, though one rather different from the ones in most consumer laptops. The screen I’m staring at as I type this is backlit with white light – it’s what’s called a transmissive display. To produce colors on the screen, there are three colored filters that can be selectively applied to each pixel – these filters allow each pixel to display a huge range of colors. But these filters block a huge amount of the light a display transmits, up to 85%, which make displays extremely power-hungry.

There are other strategies for producing color from a transmissive display – one can create the illusion of colored pixels by flashing red, green and blue pixels in sequence – unfortunately, LCD displays don’t refresh quickly enough to make this technique workable. So Negroponte is using a display that puts red, green and blue pixels near each other, blurring into a single, colored pixel. The monitor will have three colored backlights, and each pixel on the monitor will have three small lenses, etched into a lenticular screen – each lens will pick up only one of the three backlights. The screen will also be able to work in a black and white, reflective mode, useful when the laptop is being used in bright light or sunlight.

While the display is going to need to be custom manufactured, Negroponte believes it will be possible to produce it at a fraction of the cost of traditional displays, and his team is already negotiating with display manufacturers in Asia to produce the product. It’s likely to be significantly less sharp than the LCD displays we’re used to, but will use far less power. Negroponte’s goal is for the machine to work on a 100:1 crank ratio – one minute of hand cranking generates sufficient power for the laptop to operate for 100 minutes.

I had several dozen questions based on my work with computers in challenging environments, and Negroponte had excellent answers to almost all of them. 12 volt power? There’s an adaptor for that. Voltage surges? Shouldn’t be a major problem given the most power draw of the machine. Cooling? The machine doesn’t have a fan since the processor is fairly slow and there’s no disk drive. The one hardware area where I wasn’t entirely satisfied with the answers had to do with mesh wireless support – some work in mesh suggests that it’s useful to have two radios per machine to provide robust backbones for sharing connectivity, and the current machine has only one radio, likely as a cost concession.

I didn’t get to see the software being designed to operate the machine, but learned a bit more about the team working on it. A small team of Red Hat engineers are customizing a Red Hat distro to the processor and hardware specifications of the machine. They’re doing some work on the GUI as well, as are Alan Kay and Seymour Papert – the total development team is about 18 people, including Kay’s students at the media lab. The machine will come with tools to encourage students to experiment with programming, including Squeak (a graphical environment for the Smalltalk programming language) and Logo. The plan is to make the software available online in a few months so that testers can bang on it and suggest features.

Localizing the software for different languages, learning styles and environments is going to require local production of software, which Negroponte appears to be planning for. Production of the hardware locally, on the other hand, is going to be “optional”… by which he seems to mean, “Some countries are going to insist on producing this lately, but it will be near impossible for them to do so at quantities that make it affordable.” Scale is clearly a major part of what will make the laptop succeed or fail – the laptop won’t be produced unless at least five countries sign up at a million laptops each. With an initial production run of 5 to 10 million laptops, the price is likely to be between $130 and $150 per unit, not including any distribution costs, marketing, or any digital content that comes pre-installed on the box. As the project scales up, the $100 per box target comes into sight.

The laptop is not “for sale” – it’s going to be available for students only, and will be distributed through the same channels that school books and uniforms are. The laptops will be the property of children, not of the school. Colin Maclay, a Berkman colleague who’d joined me for the visit, pointed out that in many countries, school books and uniforms are sold by (highly profitable) local businesses, and that losing a book contract might be a major blow for local employers. Negroponte points out that this doesn’t have to be a revenue loss – publishers could sell the electronic rights to textbooks on a per copy basis, which might make electronic textbooks even more attractive, on a revenue basis, than paper ones.

This might complicate the economics of the device. The idea is to make the laptop required equipment for all school students and price it on a basis where it replaces textbooks. In some of the developing world school systems Negroponte has investigated, textbooks cost $20 per student per year – if the laptop is sold to students (or provided by the national government) with a five-year financing option, it costs the same amount as annual textbook spending. Except that, if schools need to license intellectual property from existing publishers, the cost would certainly increase.

Talking through the structure of the OLPC initiative, Negroponte mentions that he was having a tough time figuring out who could be the CEO of the entire project. “In the startup phase, we need someone who’s halfway between Michael Dell and some great supply-chain genius. But in three years, we’re going to need someone halfway between Kofi Annan and Seymour Papert.”

The solution – run three separate groups: OLPC hardware (possibly based in Silicon Valley or Asia), OLPC software (likely based at the Media Lab) and OLPC International (based in New York, Geneva or another global hub.) It’s this third unit – the one charged with distributing, localizing and supporting the machines, and figuring out how they get used in global schools – that I’m most fascinated by.

While Negroponte has some general solutions to the interesting problems around distribution and usage, I got the sense that there hasn’t been as much detailed thinking about the on-the-ground challenges as there has been about the physical and software design of the machine. Colin wanted to know how Negroponte was navigating the complex internecene politics of working with the various ministries of a government – is this project owned by the Ministry of Education or the Ministry of Information and Communications. Negroponte explained that he was just working with heads of state, counting on them to get their teams to implement the project. Colin and I exchanged glances – we’d both worked with President Leonel Fernández of the Dominican Republic, a visionary leader who’d dedicated himself to bringing technology to the DR in 1996. His party was voted out of power in 2000 by a kleptocrat who ran on a platform of “Plantains, not PCs”, and pro-IT reforms were rolled back. (Fernandez is back in office now, after last year’s elections.) I’m guessing that a OLPC project started under one leader could stall under a future leader.

My questions largely had to do with how the laptop would be used in the classroom. I made the mistake of asking a question of how the laptop would be used as “a teaching tool”… like Papert, Negroponte’s a big believer that students simply need access to technology and can use it to teach each other and to make discoveries themselves. When I expressed some skepticism about teachers’ willingness to use the computers in the classroom, he referenced Maine governor Angus King’s initiative to bring computers into middle school classrooms throughout the state. Initially unpopular with teachers, the laptop project is now widely viewed as a success and is being replicated in other states. It’s clear that the strategy behind the device is a trojan horse one – sell the device as an e-book, then see what students are able to do with a flexible, net-connected, programmable tool.

After peppering Negroponte with two hours of questions, I’m fairly convinced that this laptop won’t suffer the problems the Simputer did – I believe it will get produced and distributed and that the software will enable e-books, web browsing, word processing and programming. As much as I enjoy the geekery of challenging Negroponte and others on the fine points of hardware and software design for the developing world, I’m convinced that some extremely smart people are working very hard on the hardware and software side of things. While I might question some of the decisions made, I don’t know that my second-guessing is helpful at this point.

On the third and fourth fronts of the project – the marketing, distribution and maintenance of these devices and their connection to the Internet, and their use in the classroom – I think there’s a lot of unanswered questions and I think the global community of folks interested in IT in education, especially in IT in the developing world, could assist Negroponte and team with their thinking.

Specifically, I think it would be great for the OLPC team to have a set of requirements and suggestions for nations participating in the program on how to distribute, link, support and teach with the laptops. It sounds like Negroponte would like to make it a requirement that every student in a classroom has a laptop. Should it be a requirement that schools implementing laptops have internet connectivity? Can this connectivity be used the way it is in the SchoolNet Namibia project, to let schools become ISPs, using revenue to subsidize the net connection and, perhaps, the laptops? Will businesses repair the laptops? Or will students do it informally, or start their own businesses?

Colin and I are talking about soliciting suggestions on the distribution and use questions surrounding the One Laptop Per Child project and compiling them into an advisory paper for Negroponte and crew. (If you’ve got questions or suggestions, posting them on this blog is a great way to start a discussion…)

One Laptop Per Child is an amazingly ambitious and radical project. If it succeeds, it will radically change how the world learns, communicates and interacts over the next couple of decades. And if it fails, it will likely scare off anyone from trying anything this radical in technology and education for many years to come. For that reason alone, I’d like to make sure it doesn’t fail, and would love your help in figuring out ways to make sure it succeeds.