WITH THE G8 summit and Live 8 dominating the news, the divide between advanced economies and the developing world is attracting unprecedented attention.
While the headlines understandably concentrate on the fundamentals of food, water and medical care, some engineers are looking at the longer-term problem of how to bring developing economies into the digital communications era.
The internet is a vital tool for education and trade. It currently connects 100 million computers but less than two per cent of the world’s population. The reasons for this are simple. In India, for example, the average rural wage is £20 a month, putting even the cheapest PC well out of the reach of communities.
Meanwhile, in much of Asia and Africa, outside large cities power supplies and communications infrastructure are at best erratic. Over the past five years a number of initiatives have promised a solution in the shape of low-cost computing systems. Rather than creating hi-tech systems where performance is the priority, engineers have approached the issue from an entirely different angle, aiming to reduce costs, power consumption and reliance on the availability of electricity supplies.
Ideas over how to achieve widespread use have split into two camps. Some believe that portable devices are the key for developing countries, while others have concentrated on networked devices. In 2001 Indian scientists announced the development of the Simputer, standing for ‘simple, inexpensive multilingual computer’. The first of several solutions, this handheld device aimed to bring computing to remote Indian villages. But four years on, orders from government agencies and not-for-profit organisations are few, with most coming from businesses and city dwellers. Low sales mean manufacturing costs have remained high while standard PC prices have fallen, and there is now little difference between their costs.
In April, with sales at around 2,500 units, far short of the expected 50,000, Simputer manufacturer Bharat Electronics announced it was to rework the handheld’s design using low-cost chips in the hope of cutting its price to around £100. The new model should be available by October.
However, the arrival of even lower-cost systems is imminent. Earlier this year MIT Media Lab’s Prof Nicholas Negroponte announced the development of a £50 full-colour laptop marketed at ministries of education in developing nations. His aim is for the devices to be distributed to children, who can take them home after school, engaging whole families in technology. Use of an electronic-ink screen will cut costs. Made from a polymer roll, this uses electricity only when the image on the screen changes, massively prolonging battery life. However, orders will be limited to a minimum of one million units. The first should be ready for shipment by early 2007.
Some charities argue, however, that the reality of life in developing countries means that portable devices are unsuitable for local needs. Instead, ‘thin client’ systems should be used. These consist of a number of simple, pared-down computers connected to a central server.
One advantage of this is that older computers can be placed at the user’s end, as the server deals with most of the information processing. David Sogan is founder of Digital Links, a charity that has distributed almost 15,000 second-hand computers at very low cost to schools, hospitals and communities in 15 countries across Africa and eastern Europe since 2002. Digital Links is examining the thin client model in partnership with charity organisation the Shuttleworth Foundation.
A 20-school trial is currently running in the Limpopo region of South Africa. According to Sogan, business plans such as Negroponte’s, which are built on the expectation of receiving large orders, are unrealistic.
‘A country like South Africa could possibly manage this, but at the moment its biggest initiative is rolling out 30,000 computers to schools. An order of a million is a bit of a pipe dream,’ he claimed.
Laptops are also easy to steal, unlike networked devices. Meanwhile, in a harsh climate, machines must be robust — particularly in places where support and spare parts are hard to come by. Users often, therefore, prefer well-known brands such as Compaq and Dell as they are perceived as more reliable.
‘People like to use an old computer with full functionality rather than a pared-down machine. They don’t want to be seen to be using something that is different from that used in the West,’ claimed Sogan.
Sogan is convinced that power supply is the key to providing computing for all. ‘For every extra watt required the price rises substantially. A solar-powered system can cost twice as much as the PC, but outside towns mains electric is not reliable,’ he said. ‘Ideally, you would remove unnecessary power consumption to reduce costs. A computer without a fan and with a really good heat sink would be ideal.’
The British-designed Nivo could provide a solution. This uses very little power and will be available at under half the price of Negroponte’s laptop. Creator Prof John Naughton of the Open University is director of the not-for-profit Ndiyo Project — named after the Swahili word for yes.
The group has redesigned the network to create an ultra-thin client architecture that is both cheaper and more energy efficient. ‘We shrank the PC to the size of a bar of soap, running on cost-free open source software,’ he said. ‘The next step will be to make this a chip that can be inserted in the back of a monitor.’ Whereas a PC typically uses 100W of power, the Nivo box uses 5W. It has ports for ethernet, power, keyboard, mouse and a monitor.
Within the next 18 months Naughton hopes to be able to reduce the cost of each box to less than £15, a price similar to that of secondhand systems but with the advantage of using new rather than refurbished parts for increased reliability. However, the problem of providing a cheap and reliable power supply remains. With prices of both portable and networked systems falling to a more realistic level, engineers must provide the means to run their systems.
Unless a solution is found, the divide will remain.