While GIS systems continue to proliferate in developed countries, the technology is also being deployed in developing countries. However, the technological environment of such locations must be considered. For example, providing support via the internet or distributing metadata through web pages is not practical in a locale where telephone service is erratic. Likewise, training end-users in database products and office productivity applications should be approached differently when it represents the students’ first experience with a computer.
Nigeria is the most populous country in Africa, and Lagos is its largest city (see figure 1). Lagos served as the capital until 1991, when the Federal government moved to Abuja, in part, to help alleviate the stress on the population expansion of Lagos. No official estimate is available, but it is believed that, as of 1996, between 10 and 12 million people currently inhabit the city (an estimate of 3.8 million was given in 1980). The increase in population has placed a tremendous burden on the infrastructure of the city, where such infrastructure exists. Basic utilities often are unavailable or, at best, are unreliable in nature.
Amid what would appear to an outsider as a chaotic situation, local government continues to function, although hampered by a lack of funds and an increase in workload to accompany the population growth. The Ministry of Lands and Housing within the Lagos State Government is responsible for all land administration functions, including (from Di Zitti, et. al.):
With the help of a loan from the World Bank, the Ministry is implementing a Land Related Information System (LRIS), capable of providing basic GIS functions for the various Directorates under the Ministry. The first phase of the project is due to be completed in August, 1997, and will cover a 40 square km area in the heart of Lagos, but will include small scale mapping and orthophotographs covering the entire 500 square km area of metropolitan Lagos. (see figure 2).
The LRIS system provides a database of all land transaction documents along with a cadastral geographic database. A primary application will be the generation of invoices for annual "ground rent" for government-owned lands and "neighborhood improvement charges" for privately owned lands. The current paper-based system is unable to adequately determine the proper amounts for such invoices, and as a result, the ministry does not collect nearly enough funds to support its operation (current estimates show that the government receives less than 25% of the amount it should in ground rents).
In order to properly populate the database tables, the primary contractor (Reid Crowther) was given access to the vault containing some of the relevant documents. However, these documents had been glued into binders that could not leave the vault. A photocopier was placed in the vault and copies were then scanned into electronic format and sent off-site for data entry. The investigative thread for each property was often painstakingly slow, as described in FitzGerald, et. al.
Survey plans were also copied and sent off-site for coordinate entry into a CAD system. These drawings were then converted to ArcInfo format using a unique identifier.
In addition to digital mapping of a property layer and orthophotographs, geographic data was developed for street center-lines, water utilities (hydrants and mains), land zoning areas, and the government acquisition plan. These themes provide the initial base for the LRIS system. End-user applications were defined as:
Because Survey Plans from privately owned areas often contained conflicting information, a super-set polygon coverage (SUPERNET) was created for the entire project area, with each polygon containing a unique identifier. A lookup table relates the polygon identifiers with property identifiers, as shown in figure 3. Property identifiers are used along with document identifiers in a Registry search application to aid in the process of tracking ownership and transacting transfers of property.
All attribute data except the SUPERNET identifier are kept in Oracle tables, and several applications have been written using Oracle FORMS where geographic information is not necessary.
However, each application was designed to be run without the benefit of a network. While the maintenance of the geographic data occurs in a single location, other end-user applications are needed in the main State government office complex (the Secretariat) located 2 km away. In this particular region of Lagos, telephone service has not been available since the NITEL (Nigerian Telephone system) exchange burned down in 1996, and creating a private WAN using radio frequency or other technologies was determined to be too expensive. As an alternative, CD-ROMs are created each week at the Survey department’s facility and transferred to the other remote locations. A "data update application" is then required to re-load the data tables and coverages to the hard disk of the remote workstations.
The introduction of computer-based applications into business processes where such structure has been absent presents its own set of challenges beyond those of technical viability. However, the long-term prospects for such systems will often rely on one or more individuals who are willing to take some risks to ensure that the systems are used and maintained. The presence (or absence) of these "project champions" will likely dictate the success (or failure) of the project. This is just as true when implementing such systems in developed countries.
