India, Kenya, Brazil, Chile. Coffee growers, rural farmers, wine producers, village self-help groups, and rural schools. HCI researchers are used to working with diverse sets of people, but over the past two years, the UK’s “Bridging the Global Digital Divide” teams have grappled with vast distances—distances marked in miles but also distinguished by differences in social and cultural understandings, wealth, and access to information technology. These projects, funded by the UK’s science research council, are looking at new interaction technologies and design practices for those on the wrong side of the divide—those billions of users whose lives could be transformed by the advances in computing and communication technologies that we take for granted.
Each of four projects identified a developing world user community to co-design prototypes and practices:
- Fairtracing is looking at ways to link UK consumers with coffee producers in India and wine cooperatives in Chile. In one prototype, you point your mobile phone camera at the barcode of a supermarket product and receive information on the groups of people involved in getting it to the shelf (see sidebar, “Fairtracing – Simple Concept, Deep Consequences”).
- StoryBank uses mobile phones in an Indian village setting. Participants are exploring how the user-generated content revolution might play out in a context where there is low computer and textural literacy and minimal access to the Internet.
- Village e-Science is deploying toolkits of gadgets to improve both education and agricultural practices in Kenya (see sidebar, “The Village e-Science Experience”).
- Rural e-Services is expanding understanding of participatory design by working with a farmers’ cooperative in Sironj, Madhya Pradesh, India.
Each of these projects produces a wide range of findings (go to www.bgdd.org), but what broader questions does the work raise for the role of HCI professionals in emerging world contexts? What more general lessons are being learned? Perhaps you are thinking about getting involved in similar collaborations; how might the experiences shape the way you approach your design or research practices? We’ll try to outline some starting points.
First, if you are based in a country like the USA or the UK, how can you possibly justify this sort of work to yourself, your boss, or your sponsor? Is it right that a group of people in the UK run projects in countries so far away? What about the costs, the money needed, and environmental impact of all that travel, for instance? Shouldn’t local researchers and practitioners be doing this work?
While each of these UK-based teams have made several field trips, people employed from within the communities do the day-to-day work. This approach has been an effective way of building relationships while reducing costs and equipping local people for future design and development work. As for undoing the effect of long distance plane travel, you should consider building carbon offsets into the budget.
There are costs, but benefits too. Big companies, like Nokia and Microsoft, have long understood the importance of working in emerging world contexts, and for our UK team members, the insights gained more than compensate for the travel and management costs.
We are uncovering interesting interaction approaches with broad appeal—not just for developing world contexts. Take the media-sharing practices in the StoryBank project. While most of us read and interact with content like Flickr photos or YouTube videos while sitting alone in front of our computers, the villagers have showed the value of group enjoyment of content through a shared, large, touch-screen display situated in the village community center. Perhaps these experiences point to future, wider-use of such technologies in homes and other shared spaces throughout the world.
If you’ve convinced yourself that working in these places will be valuable, what sort of technologies and design practices should be on your palette as you sketch out your work proposals? Skim through HCI and development articles and you will often come across emphases on mobile technologies, particularly phones. Although we had read about the overwhelming eagerness of people in developing countries to get their hands on a mobile phone, going there ourselves and witnessing the accelerating uptake of such gadgets has been profoundly affecting. In one community, when we started work there were only a few mobile phones since there was no cell-tower coverage. After the cloud of wireless coverage arrived, within nine months the majority of villagers had access to a handset, despite the relative cost and the lack of other utilities such as uninterrupted power supplies.
The mobile phone is undoubtedly a transformative technology for development work. Networking and power management innovations, along with large-scale investment, means that even very remote, rural locations are getting connected. There’s a great need, though, for bold efforts in interface and interaction design to go beyond voice calling. We’ve seen that you cannot simply deploy standard interfaces and applications for populations that do not have our exposure to computing or, indeed, the levels of textual literacy we assume. It is vital, too, that in the rush to exploit mobile devices, we do not overlook other information technologies. Radio, television, and paper are three other highly pervasive parts of the environment that can be used in conjunction with newer technologies. Standard desktop or laptop computers are in very short supply. In our projects, this scarcity has been very challenging at times; how can you create user-generated content when everything from capturing the audio, images, and video, to the editing and sharing of the material has to be done on a small screen device, and when you can’t rely upon your users understanding conventions like menus or reading textual options?
Building systems in such a way that people can appropriate, subvert, or mash them up has become a popular theme in HCI thinking. This sort of approach might be particularly powerful in the contexts we’ve been looking at for at least three reasons:
- Designing a device or service so it accommodates a range of needs is valuable where access to such resources is limited.
- There is going to be a scarcity of professional designers and developers for a long time.
- Even experienced professionals might find it hard to keep on top of the complexity, variety, and rapidly evolving personal and community information needs.
In the Village e-Science work, for example, kits of gadgets are designed to promote a building-block type use, configured differently by groups in the community for anticipated and unexpected needs.
What about the fundamental design approaches needed in these types of projects? How should you see your involvement with a community? Do you need to re-skill? User participation in technology design is now often taken for granted. There are a number of popular textbooks commonly including descriptions of key methods like ethnographic inquiries, scenario development, cultural probes, and paper-prototyping. In developing world contexts, however, the meaning and impact of participation must be thought about carefully. What sort of “participation” by local communities is “right”? We found a useful framework in Projects with People: The Practice of Participation in Rural Development by Peter Oakley and his collaborators, who distinguish three types of participation:
- In participation as contribution the beneficiaries (them) are invited to contribute to a project that an external group (us) has initiated for their benefit (as perceived by us, the outsiders); we determine the timing and form of the participation. Such an approach reinforces attitudes of dependency and cannot promote significant social change. Mainstream participation in technology design—in the developed world, that is—perhaps has too much of this flavor.
- In participation as organization, the local community is empowered to establish social structures to support their needs. This approach can be valuable, but often involves a subtle (or not so subtle) cultural colonialism.
- In contrast, participation as empowering and leading social inclusion demands that participants are engaged in defining project goals, in negotiating activities and methods, building up their control over the process, and gaining confidence and skills in asserting their own objectives.
As we’ve already noted, realizing this high level of participation leads to extended fieldwork. It also means letting go, understanding that our co-designers, and not we, the “experts,” are in control. In the Indian Rural e-Services project, for example, we have had to adapt software cycles to fold into the agricultural rhythms that dominate our participants’ livelihoods; you can’t do a brainstorming workshop when crops need to be gathered.
To sum up, stepping into development work is costly, but there are ways of managing the complexities and distances. You will be challenged to see familiar technologies and design practices in bright new lights. If you are a graduate student just finishing your HCI studies, you could do far worse than to take a year and engage with the types of rich communities with whom we have partnered. If you are an experienced researcher, go and review your methods and your take on future technology. Or if you are company-based designer, persuade your management that co-designing with representatives of the billions of underserved users of the world makes absolute market sense. Whoever you are, whatever you do, get involved and shape a better world for us all.
Fairtracing – Simple Concept, Deep Consequences
Fairtracing is exploring tools for connecting developing world coffee producers to developed world consumers. What we thought would essentially be an educational tool for consumers about the value chain that led to the products they buy, with perhaps some useful feedback from consumers to producers, is rather more.
Interviews with producers revealed that they too will learn about the journey of their products, particularly in cases where this information isn’t currently obvious. If a group of growers find out that subsidies are being paid further up the line, this information equips them to bargain. If they trace how a myriad of small traders add their percentage to the cost, this might goad them to improve upon distribution channels. Of course, as the tool gets used in this way, the content provided by people along the value chain might well evolve. We do not, for instance, expect interim traders to offer up information willingly that would prejudice their livelihood.
This is just one example of interface-impact our design process has to accommodate. An audit of politically sensitive information types yields a range of concerns that we have to consider, issues that go way beyond any cognitive or aesthetic considerations:
- Balancing effective marketing of products with consumer interests.
- Representing the cultural and social identity of value chain actors in a way that is intelligible across cultures.
- Expressing the relationship between the developed and developing world respectfully to all interests.
- Displaying ethical and environmental commitments fairly, both absolutely and relatively, in the context of the production conditions and national legislation.
- Testing environmental information against consumer trends. (Our original prototype map interface has been adapted to make food miles an issue, reflecting changes in the representation of environmental issues, whereas a different representation might make the lower carbon footprint of production in warmer climates more apparent.)
- Showing reflexive concerns such as the transparency of who authored what information and their relationship to the chain, so it is clear whether they are promoting the product, or commenting on it.
- Showing gaps non-prejudicially, given that not all value chain members will contribute and that some data won’t be available across all products.
- Giving a coherent and authoritative feeling to the information (so that it is easy to assimilate) while acknowledging that it is user-generated and may be qualitatively very different between products, producers, points in the chain, and so on.
The Village e-Science Experience
Sub-Saharan Africa is the challenging environment where Village e-Science (or Vesel) aims to bring information and communication technology (ICT) to farmers by involving the whole community, including schools and the community radio station.
Our experience of the context was patchy since we came from geographically distributed UK universities and were an interdisciplinary team that included HCI specialists, educational technologists, radio engineers, electrical engineers, and sensor network specialists. Fortunately, we formed a partnership with the University of Nairobi’s specialists in ICT and agriculture, which meant that they could deploy the ICT resource kit in the villages and reduce the need for field visits by the UK teams. Few buildings in Kambu have access to electricity, and participants learned to use a portable solar panel with a storage battery to power an Apple laptop, iPod, and cameras.
The project involves two contrasting communities in Kenya. The first is in a lush area with small market gardens, where the farmers need to market crops for local and EU markets. The second has large plots and drought problems that need support with crop selection and irrigation; it also has the only local community radio station in Kenya. We had learned that in this physical environment, mobile phones and radio, rather than desktops that would be costly in terms of energy and maintenance, would be key in interactive education and communication.
We discovered that the portfolio of tools and techniques for participatory design we arrived in Kambu with were not adequate or appropriate for remote rural users, some with low literacy, no ICT experience, or even access to electricity. One of the main concepts we have used in redesigning our methods is Ubuntu, a southern African traditional worldview that stresses collective contribution, solidarity, acceptance, dignity, stewardship, compassion and care, hospitality, and legitimacy. The word Ubuntu comes from a Xhosa expression meaning, “a person is a person through other persons.” For example, low-fidelity prototyping and evaluation methods based on the verbal protocol seemed very unnatural to the users and did not work. Techniques that have worked include cultural probes, the use of cameras to capture user stories, picture sorting, and work shadowing. We used all of these to gather initial requirements and then used Rich Pictures for scenario-based design.
We learned the importance of getting the backing of the whole community, not just the community-selected key users who would learn to use the applications before training groups of others. We learned the value of working with small groups of users for collective exploration and evaluation. Identifying key individuals within communities is vital, as this extract from the notes of one of our local researchers, Souleymane Camara, highlights:
“Today I was training Radio M with their website. But invited teachers from Silanga school to come along if they need for a couple of more training. Rose is the lady in charge of the kit. She is so determined to master it, that she calls the attention of anybody who is around and has a few skills to offer when I am busy with other trainees. My assistant Peter from the University of Nairobi did not have a rest from her and John, another IT-literate teacher. It is impressive to see how she does not let her colleagues dominate the kit as she tries to get to the same level. She is still in the office right now and it is 6:47 p.m. Every time I come around to show her something she is already on another level (browsing BBC website for schools, sending a post to the blog, changing the desktop background and other properties of the computer).”