Thursday, November 20, 2008

FADECO: A Model To Follow

This item was found in the “AfricaFocus Bulletin” and is a reprint of a portion of the article: Unbounded Possibilities: Observations on sustaining rural information and communication technology (ICT) in Africa Written by Ian Douglas Howard on behalf of the Association for Progressive Communications (APC) - October 2008. References to “the author” in the text are to Ian Howard. Also, the text of the article uses British common usage of English and that that commonly used in the U.S.


Case Study #1: Grassroots ICT Development in Tanzania - FADECO

8.1 Background


FADECO is located in the small town of Karagwe in northwestern Tanzania near the Burundi and Ugandan borders. The town, marked by a police outpost, a few schools and a concentration of mud and some cinder block buildings, rests on a table-top plateau that overlooks lush valleys where bananas, coffee, and staple foods are grown. The road from Bukoba, the regional capital, is unpaved - like most roads in the large east African country. The drive takes several hours, weaving past farmers carrying their goods on their heads, bicycles, scooters and the occasional car, all cloaked inred clouds of dust along the winding and bumpy road.

Registered not-for-profit NGO founded in 1996 by a group led by current director Mr. Sekiku Joseph Mtabazi, FADECO is a small and very modest association. In addition to the director, its principal staff consists of Mr. Itegereize Titus Tobias (chair), Mrs. Elieth Kikaka, (office manager) and Mr. John Kibuuka (information technology manager). The organisation works to provide information resources that help families to improve their living standards. It serves as a vehicle for promoting new agricultural methods and other activities to heighten community and economic development.

Where possible these efforts have been commercialised as separate ventures. For example, Sekiku began a very small seasonal fruit drying business, FADECO Trading Co. Ltd., based on techniques that he promoted via FADECO. As well as acting as chair, Mr. Titus sells agribusiness products that support the farming techniques taught by FADECO, such as solar drying and composting. FADECO largely serves as a brand name for initiatives promoted by the group, as it has few resources.

The group has maintained an apolitical nature, partly to avoid any confrontation. As noted by Sekiku, the organisation allows the group to participate in not-for-profit initiatives where there is funding and no commercial interest. The association is best described as a manifestation of Sekiku's interests and a body that legitimises his endeavours.

In 1997, Sekiku, a self-taught technologist, began work on a small telecentre for the community under the FADECO umbrella. This centre was based on his property in the building next to his home that was previously used for the aforementioned fruit-drying business. Sekiku purchased used computers for a few hundred dollars each ... They were then connected to the internet via modems on a fixed-line telephone network. To pay for some of this equipment, Sekiku received small grants and donations from the Dutch NGO the Humanist Institute for Development Cooperation (Hivos), the British Council and others.

In 2004, after a few years of operation, he was able to buy a VSAT (a weighty purchase at more than USD 3,500) with the financial support of an NGO called the Regional Agricultural Information Network (RAIN). He purchased the VSAT so that he could avoid the onerous charges for internet access. To access the internet, his telecentre previously had to dial out to Dar Es Salaam, 1,500 kilometres away, with fees for these long-distance calls calculated based on distance. ,,,

In late 2006, ... Sekiku began to build a wireless network. His intent was to share the internet costs with other groups in Karagwe. He started his network with a few off-the-shelf wireless access points and at each site used directional antennas to point back to his base station at FADECO. In August of 2007, his network connected three clients: 1) a private secondary school; 2) the local office for the electric company (Tanesco); and 3) a local agricultural development NGO.

8.2 Observations

Sekiku explained the long process of discovery and frustration that he experienced while connecting these customer sites to his network. He toiled slowly, learning piece by piece how to install equipment and debug problems with help from colleagues via email and online chat, and using online references such as the WNDW books and internet forums. Eventually, however, he did persevere.

This case study demonstrates that learning is critical to allow such networks to be installed and supported. Sekiku was not taught, but learned how to build the network by searching for information and through trial-and-error. This learning process did require access to good learning materials, some counsel and reasonable access to equipment, but for the most part it was accomplished through sweat and passion. The process of taking ICT graduates from local universities and developing them into technologists can take considerable time. This process can also be counter-cultural for many, where the norm for young employees is for them to rarely be asked to make decisions on their own. This resistance to self-learning must be broken.

Opportunities for self-learning are further hindered because few of these capable minds have the opportunity to tinker with computers or other gadgets, or break them as do many techies from richer markets. Learning the troubleshooting process is central to becoming technologically minded. This analytical process can take considerable effort but can be done. The many ad-hoc roadside bicycle, car and television repair shops across developing nations are a testament that this ability exists everywhere but that it is not well cultivated by most developing nation schools. ...

In most ways the development of this network defies conventional best-practices, which recommend that ICT projects are planned, people trained, and equipment selected based on design and evaluation before commencing work. Sekiku's approach missed each of these steps, working in a piecemeal fashion with few resources and no formal training. ... and, yet, the network was built and still persists. Thus, it has succeeded in becoming sustainable.

The slow process that Sekiku undertook allowed him to build the needed skills at the same pace as the network's growth - not by design but due to access to equipment and funds. This slow development is contrary to standard practice. Generally, telecentres are built rapidly so that they can be operational as soon as possible. Building quickly is perhaps necessary where rent and salaries are high, and technical competence and equipment abundant. In this case growing slowly was not a problem and is the norm. By building slowly the staff can learn how to support the network as it grows in complexity. This contrasts with many other sites that the author has visited, where local staff members have not appropriated the skills required to support the systems or services provided by the site, resulting in their degradation.

8.2.2 ICT sites that do have readily available technical support should adopt new systems only at a pace at which their staff (or their support network) can competently learn how to use and support them.

In 2004, the author visited the University of Bamako at the behest of USAID, which had sponsored a project to install a wireless network there. Although the system was reasonably well installed by local and foreign contractors, it fell into disrepair within a few months. Funding for continued technical support had ended and local staff had limited knowledge of how to support the system. Thus, many of the sites were completely offline, while the network at large was so overwhelmed by viruses that it was virtually unusable. Staff had been provided training, but most had little prior technical experience and very little experience with networking.

Some believe that a techie can be made by sending someone to a five-week training program. The author has found, however, that the attributes required of techies (such as the ability to persist, self-learn and troubleshoot and profound curiosity about how things work) must already exist as they cannot easily be taught, but generally only fostered. This site reinforced this notion. Even after many weeks of comprehensive training by the implementers, these former administrators were evidently not transformed into technicians.

When the author's team arrived at the university, his team set about a slow process of teaching a few of the staff how to repair these problems themselves. They identified those staff members who possessed techie attributes and gave them responsibilities. Eventually, with a lot of guidance and on-the-job training, they were able to repair the network themselves. A little more than a year later, the university's staff independently and successfully moved several nodes and installed some new sites.

8.2.3 The need to localise expertise and provision for local repairs and within local means becomes more necessary with increasing distance from major urban areas.
The ability of local staff to support their network increases in importance as locations become more remote. In isolated places like Karagwe there are great costs to bring in expertise or equipment from commercial centres such as Dar Es Salaam, Nairobi, or Kampala.

8.2.4 Commercial grade wireless systems are often cheaper once training and long-term support is factored into total costs because of the inaccessibility and lack of local resources in rural developing communities. There is also a paradox in the use of low-cost wireless/ICT systems in the developing world. The factors that make this equipment relatively cheap in the developed world are less relevant in rural developing communities, where affordable equipment is generally of poorer quality and very basic. Low-cost wireless routers are designed to sit in the corner of a climate controlled office and connect a couple of laptops nearby. When these low-cost routers are subjected to the heat, humidity, and/or dust of many developing countries, they quickly fail.

When there is a lack of competent technical staff in rural developing communities, installing poorer grade equipment can raise support and training costs and increase risks of failure. This is often overlooked by western project proponents who see low-cost tools as a way to overcome limited resources and do not consider the costs of supporting such systems - costs which are largely absent in developed world cities. ...

8.3 Financial Analysis

Money flows through in small amounts and seasonally nourishes this business, not unlike the farms and other enterprises that surround it. Appendix one shows the FADECO telecentre's income statement for a typical month in 2007. As the bottom line of this table indicates, this business loses money, but it continues to persist because it is sustained by its operator, Sekiku.

The business model that has emerged for this site is one that generates revenue by providing a variety of ICT services including training, internet use, and wireless internet access to the three customer sites. There is not a formal business plan, nor a business per say, so this assertion is based on facts gathered on-site. The telecentre has also begun a FM radio station which supports itself through announcements, advertising and sponsored programs. Though not an internet service, the radio is an inexpensive way for Sekiku to extend the internet out to the poor, those who cannot read, and/or are not computer literate by reading what they find on the internet over the air.

Typically the telecentre earns about USD 370 per month. Costs can be broken down into operating and staffing costs. Operating costs are paid to service providers and amount to an average of USD 400 per month. Staffing costs amount to almost USD 340 per month. However, this figure indicates only what is paid in ideal circumstances, and staff members are typically only paid when there is sufficient money. Most often, they only take home only a fraction of this amount. Staff members are quasi-volunteers, with no formal status. They could best be described as "casual wage workers," meaning that their wage is not defined strictly, but they are paid according to the resources available to the telecentre.

While this arrangement allows FADECO to continue operating despite little revenue, it means that staff members treat their work as a hobby that they attend to after taking care of family and other personal interests. It limits the professionalism of the business ... [nevertheless] This flexibility in staffing is perhaps why FADECO and many, if not most, businesses can survive in such rural economies with wild variability in incomes due to seasonality and commodity price fluctuation.

Costs are also scaled in other areas, as with telephone use. Because neither the telecentre nor Sekiku have immediate access to credit, expenses are naturally scaled back when there is no cash to buy phone credit. These scalable costs allow the centre to adjust to fluctuations, though they make it quite difficult to plan or keep consistent service levels.

Beyond the telecentre, Sekiku is able to do ICT consulting work based on the expertise he developed through the telecentre. Mostly derived through conferences, consulting is his greatest source of revenue and nearly pure profit. In 2007, he earned about USD 6,000, up from USD 4,000 in 2006. Through these revenues Sekiku subsidises the telecentre, contributing approximately USD 200 per month to pay its staff, buy equipment to expand its services, and repay the three-year loan for the building. This, while the telecentre is a loss-making venture, it does provide Sekiku the opportunity to learn and gain income through consulting work


FADECO

Association for Progressive Communications

AfricaFocus