You’ve likely heard a lot about efforts to bridge the digital divide so students from families that can’t afford a home computer can at least have access to computers in school or at the local library. But in many developing nations, that particular gulf is of little concern when the three Rs of a basic education are not even guaranteed because both electricity and teachers are in short supply.
IEEE Fellow Krishna Palem, a computer science professor at Rice University, in Houston, heads a group developing a device intended to help children in poor countries learn despite these daunting problems. Focusing on students in rural southern India—most of whom have never seen a computer and for whom a small black slate chalkboard represents classroom technology—Palem nad his team of researchers are developing a tablet computer–like electronic device that will be inexpensive, easy to use, and able to operate without the region’s creaky (and sometimes nonexistent) electric power grid.
At the media event celebrating the IEEE’s 125th anniversary last March in New York City, which featured presentations by IEEE members working on breakthrough technologies, Palem introduced his device, called the i-Slate. Roughly the size of a sheet of paper, it has an 18-cm LED screen and weighs just under a kilogram. And perhaps best of all, it is powered by photovoltaic cells, so it will be grid-independent. The i-Slate will let elementary school students practice subjects such as math and shape recognition by themselves and without a teacher.
The machine does not run an operating system and is therefore not considered a computer. It doesn’t need one, Palem explains, because it’s a dedicated device, performing only one task at a time—unlike a desktop, which can do several things simultaneously. The i-Slate will be preloaded with a curriculum being developed by an educational partnership. This group is made up of researchers at Rice University’s Value of Information–Based Sustainable Embedded Nanocomputing Center; the Institute for Sustainable Nanoelectronics at the Nanyang Technological University, in Singapore; and the International Institute for Information Technology, in Hyderabad, India. They hope to have a more advanced prototype than the one Palem showed off at the March media event ready by February 2010 and a final production device by the end of that year.
Palem says the target price for the i-Slate is cheap—less than US $40. This is possible, he notes, because it will run on sunlight alone.
EDUCATED GUESSES During his IEEE media event presentation, Palem described a class of low-power chips, called probabilistic complementary metal-oxide semiconductors, that lets the i-Slate designers make a trade-off between accuracy and low power. The probabilistic chips, under development for more than three years, arose out of the realization that in audio or video applications the mental processing behind the human eye or ear makes up for glitches in what is actually seen or heard. This means that chips don’t have to run at the high-power settings that keep the signal-to-noise ratio high. In the i-Slate, where the LED screen will present simple mathematical statements such as “5 + 3 = __,” the precision that most computer makers demand in their chips is unnecessary. Other possible applications for such chips include hearing aids, cell phones, and the audio- and video-processing portions of the electronics in such devices as digital still cameras and handheld video cameras.
Letting chips perform estimations instead of down-to-the-smallest-decimal calculations yields two important benefits for the i-Slate. Its chips draw less than a watt, roughly half the power of conventional chips. This makes it possible to run the device from the energy provided by tiny solar panels shoehorned around the screen. The sunlight shining through a schoolhouse window would be enough to power the machine up. The other, equally important benefit is that chips with lower accuracy cost less than conventional ones.
THE BUSINESS SIDE Palem’s group is charting a different course than the one taken by the One Laptop Per Child initiative, which was focused on providing a $100 computer to students in developing countries, with government agencies acting as purchasing and distribution agents.
“Our initial plan is to work with the Villages in Development and Learning Foundation, a nongovernmental organization in India,” says Palem, who is also in discussions with an NGO in South Africa. Palem admits that reaching all of the roughly 106 million Indian students with no access to electricity and the more than 400 000 elementary schools in India that have fewer than teachers on staff may be impossible. But he says he has faith that the NGOs will use self-sustaining economic models that have worked well in developing countries in the past.