My December 2007 Editorial praising the One Laptop Per Child (OLPC) program (see
Negroponte’s Miss is a Hit) raised some very important questions from readers about whether pushing computers and Wi-Fi networks on people who don’t even have enough to eat at home was the most appropriate use of scarce aid funds. Given the hard facts that 1.2 billion of our fellow passengers on Spaceship Earth live on $1 a day or less, and 800 million people are going hungry, what level of priority should we place on getting these folks access to computers, the Internet, and the services that they deliver? I don’t have any definite answers to offer, but a talk I attended this week may provide a few of the pieces to this complex puzzle.
Dr. Marc Fiuczynski, a computer science researcher at Princeton University, didn’t set out to look at how developing nations can use the technologies he works with to lift themselves out of poverty. In fact, his interest in the intricacies of extremely large distributed computing networks would seem to have nothing to do with the modest low-power PCs that are so often used by organizations like
Inveneo and
Geek Corps to deliver computing services, Internet access, and the skills to support them into small African and Latin American towns. Fiuczynski became involved almost by accident when he started inviting African universities to join
Planet Lab, his current project to build a global infrastructure supporting the development of new network services. He quickly discovered that while several universities had hardware that would easily support a node of the global network, the cost of the bandwidth to connect them to the Internet was impossibly high: 1000 to 100,000 times what we’d pay in the US, Europe, or most parts of Asia.
Undaunted, he worked with several colleagues to adapt the technologies developed by Planet Lab to build local distributed computing systems that use lower-cost inter-campus connections, local infrastructure and low-cost Wi-Fi links. Besides helping bring up local Planet Lab networks in several countries, Fiuczynski is working on advanced web caching techniques that can efficiently identify and store Terabytes of frequently-used web content locally to help make the best use of whatever Internet bandwidth is available. Another part of his efforts is devoted to building systems and protocols that can cope with the much larger levels of loss and latency that occur in these networks. For more information on these robust networks, download his
presentation ROADS (Real Overlays And Distributed Systems) for Developing Nations.
While many of these projects will take years to yield tangible results that will make a meaningful impact on the economies of Africa or Latin America, we’re already beginning to see several spin-offs that are yielding more immediate benefits. For example, the same technology is already being used to develop a more robust kind of Wi-Fi router/server that has enough processing power and storage capability to support a VoIP network and delivery of educational services on low-cost computers like the OLPC and
Intel Classmate. During a recent trip to Brazil, Dr Fiuczynski got to see one of the first OLPC deployments and how it’s actually enhancing education for both students and their parents. Besides dramatically improving school attendance and providing some level of personalized electronic tutoring, the small laptops offset a large percentage of their cost by enabling distribution of electronic text books and other educational media at nearly no extra cost. Fiuczynski does not feel that either the OLPC or its competing Classmate is a perfect solution and expects the next generation of educational PCs to look like something more like an iPhone, but even the simple first-generation machines in the field today seem to be helping kids learn the skills they will need to get better-paying jobs or start small businesses.
Fiuczynski’s travels have also allowed him to see first hand how adding a little networking technology to a developing economy can supercharge the traditional mix of educational, agricultural, and health care aid. For example, the
Aravind Eye Clinic is using a
telemedicine network in India that enables clinicians to examine patients in remote villages, prescribe simple treatments that can be obtained locally wherever possible, and only bring the cases needing more intensive attention to a bigger clinic or hospital. Fiuczynski is helping develop a similar network in Belize that uses moderately-skilled field workers and low-cost laptop-based ultrasound diagnostic systems to collect images and transmit them to specialists for diagnosis over long-range Wi-Fi networks. His preliminary analysis indicates that besides delivering health care to places it would not normally ever reach, it allows a doctor or medical technician to serve 10x – 20x more people than physically making the rounds to a series of traditional remote clinics would.
Despite all these success stories, Fiuczynski says that the most important lesson he has learned is to listen closely to the people he’s working with to understand what they really need. It’s no surprise that many of the best applications of computer technology he’s working with are based on the requests from the local folks, who have an intimate knowledge of their country’s social and economic dynamics. His African colleagues have also discovered how important listening is. The Information Communication Technology Policy
section of Africa’s Technology Policy Studies Network is looking for clues to shaping their own educational and technology policies in the success stories of countries like India and Ireland that have used computing and networking technologies to help their economies become more productive and move them into the global marketplace.
While food and medical aid, agricultural assistance, education and infrastructure development are still critical parts of easing the misery on developing nations, it looks like appropriate amounts of computing and communication technology can greatly enhance their effectiveness. Questions still remain about which technologies and deployment strategies are going to yield the best results. The answers we get in the next decade or two will also be valuable to more developed economies as we all try to develop new strategies for generating sustainable prosperity in a more tightly-connected global economy.
Comments? Questions? Thoughts on bread versus bandwidth? Write me at
lhg at en-genius dot net or post your comments on our blog.
