DOHA, Qatar — April 17, 2009 — When Microsoft Research’s India lab was launched in 2005, assistant managing director Kentaro Toyama found himself far from his work in Redmond, Wash., where he had been researching computer vision and other highly technical projects.
Instead of pushing the boundaries of technology itself, part of his new charter at Microsoft Research India was to investigate the intersection of technology and society with India as a test bed. Toyama soon realized the importance and scale of the challenge.
“In my first trip to India, I visited a number of rural villages that had telecenter projects,” Toyama says. “I realized that as an industry we understood what we could do with technology, but we didn’t have the first clue about what it means to help a rural farmer living in a remote village. We needed people who by temperament and training were interested in understanding people’s daily lives in these economies, and how they might interact with technology.”
In the years since, the business community worldwide has come to a similar conclusion. In a recent BusinessWeek article it was estimated that so-called purpose-driven entrepreneurship has become a $40 billion industry worldwide.
Toyama has overseen dozens of research projects in that time, but one of his biggest accomplishments has been to co-found the tech industry’s seminal conference in this specialized field, the International Conference on Information and Communications Technologies and Development (ICTD), which takes place this weekend in Doha, Qatar.
In four years, the conference has become noteworthy enough in the tech industry that for many the term ICTD is synonymous with work being done for developing nations. This year, keynote presenters include Bill Gates, chairman of Microsoft Corp., and Carlos A. Primo Braga, director of Economic Policy and Debt in the Poverty Reduction and Economic Management Network (PREM) at the World Bank.
In contrast with traditional technology conferences, ICTD features a broad range of sociologists, economists, anthropologists and others seeking to understand how technology and humanity can interact for the greater good.
“This field requires a mindset that is open to looking beyond the technology,” Toyama says. “Computer scientists are critical, but they need to broaden their vision to look at the full context of these problems and not just the ones that technology might narrowly address.”
The fundamental problems tackled by this group are deep and systemic. The constituencies are poor countries that generally lack financial resources both at the personal and national level. Many of the populations are undereducated and often illiterate.
“Poverty maintains itself because there is no single way out,” Toyama says. “The big challenge for us is to see whether we can bring the power of technology to help solve these problems.”
When Does IT Make Sense?
According to Toyama, simply adding technology doesn’t necessarily help.
“People suggest that if you just provide computers, eventually students will be able to pull themselves out of a bad educational setting,” he says. “But these are places where electricity is often scarce or lacking. The students themselves may already be behind the curve in terms of literacy. Teachers are usually poorly trained. All of these things just compound each other.”
As a result, researchers must find clever ways to integrate technology into a practical solution. One method is to work with a strong institution and find ways to support its goals, rather than trying to lead the charge and completely solve a problem. And of course, researchers must do a lot of work up front to fully understand each situation.
“You have to work with a strong school and then see if the laptop can amplify what that school is doing,” Toyama says. “But then, how do you get to the weaker schools? Those are kinds of challenges that the community as a whole is trying to understand.”
One project being presented by Microsoft Research at ICTD this weekend, in collaboration with researchers at the Massachusetts Institute of Technology, shows how throwing technology at a problem can lead to even more confusion.
Called “Evaluating the Accuracy of Data Collection on Mobile Phones: A Study of Forms, SMS, and Voice,” the paper effectively asks whether human services are more effective for some tasks than technology.
“There is a lot of excitement today in using mobile phones to report different kinds of data in rural settings,” says Bill Thies, a researcher at Microsoft Research India. “People are using them to report health data, environmental data, financial data and so on to centers in urban locations that in turn provide different kinds of services to rural areas.”
According to Thies, however, no one has systematically evaluated the accuracy of that data. Cell phones have small keys and a small screen, and often the people entering the data have limited experience and education.
“So we did a user study looking at some health workers in rural areas and asked them to do mock patient interviews and enter the data into the phone — patient temperature, weight, symptoms,” Thies says. “What we found was a surprise. Using interfaces that many people use, we saw error rates of about 5 percent.”
When it comes to healthcare, of course, that five percent adds up to a major problem. The alternative? Human call centers. The study compared mobile phone data entry with results from live calls placed to an operator who recorded the data.
“There we found it was very accurate, about one half of one percent, although we had a small sample size,” Thies says of the error rate. “So the lesson is to consider the option of using a live voice operator, especially in a country like India, which has a relatively good supply of affordable, educated talent in urban areas to staff those centers.”
Thies’ paper has already influenced at least one tuberculosis treatment center to switch its plans from a menu-driven, mobile phone interface to a live operator.
Helping People Help Themselves
While Thies’ work shows one way that institutions can be more adept in deploying technology, Microsoft Research India is also looking at ways to help people become more adept at using it.
Another paper at ICTD this year, titled “Kelsa+: Digital Literacy for Low-Income Office Workers,” looks at ways to help people educate themselves on technology without the expense of deploying telecenters, or public computer terminals, in rural areas, where electricity can be hard to get and tech support is nonexistent.
“Instead of trying to reach communities in a remote area, we tried it in an office in a city, where the support staff that works in the building — housekeepers, security guards, drivers — also have incomes as low as $100 a month, and despite the fact that they work next to computers all day, never have a chance to interact with them,” says Aishwarya Ratan, associate researcher at Microsoft Research India.
With Kelsa+ the research team made an Internet-connected PC available to workers in an office for use any time they were off duty. The PC wound up being used about 10 hours every day, and moreover had a noticeable effect on the digital literacy skills of those who used it.
“In a handful of cases, staff members have been effectively able to upgrade their skills and jobs and find new work at data entry companies,” Ratan says. “That will provide them with a much greater chance for advancement.”
It’s another example of how in the field of ICTD, social research stands alongside the traditional framework of technology research, where varying form factors are explored for their effectiveness at solving a given challenge.
A third study being presented at ICTD this week represents a more traditional approach, looking at the effectiveness of so-called “featherweight devices” in disseminating information to areas where traditional PCs are pretty much out of the question. The paper, “Featherweight Multimedia for Information Dissemination,” was developed in collaboration with researchers at the University of Toronto and Digital Green.
“One of the conditions in working in these poorer environments is simply that people and many organizations are nowhere near being able to afford even a $100 PC,” Toyama says. “How can we take some of the value of a PC and encapsulate it in a form that is much lower cost? Featherweight multimedia involves very low-cost audio electronics combined with paper for the graphics, allowing you to create a basic multimedia experience.”
It’s less complicated than it sounds — think of the audio tours offered by some museums, where sequential audio segments are combined with a corresponding numbering system on the exhibits. Featherweight multimedia devices can be used in much the same way to communicate registration procedures at a hospital, new agricultural techniques to farmers in the field, or to walk someone through a micro-lending application. Very basic devices can be strictly audio and paper, while more advanced applications may work via a mobile phone.
“We’ve tried this with a number of different combinations of technology and paper, some of which cost as little as 50 cents,” Toyama says. “The systems are usually fairly rich in that they allow some kind of interaction, and since they provide graphical as well as audio information, they can work well even for someone who is illiterate.”
Altogether, Microsoft researchers will present six papers at the conference covering areas as broad as agriculture, healthcare, education, microfinance and illiteracy. Most of the papers were written in collaboration with academic and industry partners.
The project behind another paper received software, hardware and financial support from Microsoft Research. “The Case for SmartTrack,” describing a project to develop an ultralow-cost, cellphone-based drug information system, was written by researchers from New York University, Columbia University and Leap of Faith Inc.
It’s Everybody’s Business
So what’s in all this for Microsoft? According to AnnaLee Saxenian, professor and dean of University of California, Berkeley’s School of Information — who co-founded ICTD along with Toyama and Carnegie Mellon University’s Raj Reddy — the one thing everyone agrees on is that the opportunity to combine technology and business with positive social change is an exciting one.
“The world has how many billion people, and there are four billion still living in poverty?” says Saxenian. “It’s not just a matter of the future of Microsoft, it’s really the future of the largest portion of humanity.”
On a basic level, that sense of purpose resonates with everyone involved in this emerging field. The culture of Microsoft is one that believes in the power of technology to make a difference, a philosophy imbued by the company’s co-founder, Bill Gates.
“Bill Gates is a big fan of this work because of his joint interest in technology and development,” Toyama says.
In fact, Gates’ keynote address at ICTD this weekend is expected to focus largely on the value of rigorous research in global development.
“He’s a great advocate because he’s so focused on real impact; he’s not interested in hyping projects that haven’t proven themselves,” Toyama says. “He has a kind of ‘skeptical optimism’ that I think is essential.”
There are several other reasons why companies of all sizes are investing in developing nations today, some of which involve business strategy, and some of which don’t. At a certain point, says Toyama, corporate citizenship is just good business.
“Microsoft is frequently asked by the governments of developing countries, as well as large multinational organizations such as the United Nations and the World Bank, how they can use technology in support of their constituents,” he says. “Being able to answer those questions is important for those organizations, for those constituents, as well as for Microsoft and its relationships around the world.
“Ultimately, what matters most to everyone concerned is the overall economic situation in the world. To the extent we can help these economies grow, we create new markets and new demand, and that helps everyone down the line.”