Raffaele Favilli
Managing Director of Mappamondo Informatica, Florence Italy, R.Favilli@fol.it

Luca Morandini
GIS Consultant, Terni Italy, l.moran@ita.flashnet.it

Abstract

Authors deal with implementing Cadastral GISes in Eastern Europe, with an emphasis on standards adoption. They draw from their experience in EU funded cadastral projects, located in Eastern Europe and former-USSR countries; hence the focus is on practical aspects of developing such peculiar GIS systems.

1. The first part of the paper reviews different architectures used in three Cadastral projects, which are: A hybrid approach combining CAD data editing, a GIS querying engine and a relational DBMS for the storage of both geometric and tabular data
2. A full topologic GIS, coupled with a relational DBMS storing tabular data
3. An universal DBMS, which stores both geometric and tabular data on a uniform persistence mechanism and provides a unique programming interface to editing and retrieving them

The second part deals with adoption of standards in a cadastral GIS, ranging from standards in data format to standards in system development.

In the final part, some conclusions are drawn about the adoption of standards in GIS systems. Authors believe that, while in the western countries the adoption of standards means providing quality to users and protecting customers’ investments; in the emerging countries of Eastern Europe and former-USSR this may foster the growth of a world-class GIS industry. This industry may leverage on well-educated GIS professionals and use the globalization of markets to provide GIS solutions worldwide.

THE ROLE OF CADASTRAL GIS IN EASTERN EUROPE

Europe is becoming one, from the Atlantic to the Russian border: this message was made clear by the selection of East European countries as candidates for European Union membership. This means that the two halves of the continent have to be harmonized in some respects: legislation, market economy, and infrastructures.
A cadastre combines all this aspects: it has a lot to do with legislation, it is a cornerstone for building a land market, it can be used for decision making about infrastructures.
To this end, the EU has funded several projects (under the umbrella of the PHARE organization) to develop cadastral systems throughout Eastern Europe. Often the GIS component of these projects is the biggest GIS project undertaken in a country; hence, it leads the way for the rest of GIS industry to follow. Moreover, a successful cadastral project provides a database which others GIS applications can use, leveraging the whole GIS sector.
Authors gained hands-on experience in three EU-funded cadastral projects, in Kazakhstan, Bulgaria, Albania; we would like to briefly show these three cases, detail some standard-related aspects we encountered and draw some conclusions about our experience in implementing standards.

APPLICATION CASES

KAZAKHSTAN: LOW COST RELATIONAL STORAGE

ORGANISATION, FUNDING, HUMAN RESOURCES, DATA SOURCES
Although Kazakhstan is not Europe, it faces the same challenges of other Eastern Europe countries, hence, authors decided to include it in this review.
The project " Establishment of a Provisional Land Information System and a Pilot Operation" was conceived in the frame of the wider project "Institutional support in the area of policy formulation and agricultural statistics" started in 1993. Specific project objective was to provide technical advice and management support for the realization of the Provisional Land Reform Information System. According to the Guideline, a pilot operation on a territory of a district of about 20,000 inhabitants was carried out. The project was funded by TACIS program on behalf of Ministry of Agriculture.

THE NEED FOR AN INEXPENSIVE SOLUTION: HOW TO COMBINE LOW-COST TOPOLOGIC EDITING AND RELATIONAL STORAGE
Authors were convinced that is not the power of computers but the organization and criteria of analysis/design that make a system tick, hence, the challenge was maximizing design phase and minimizing investment in hardware and software.
The hardware was limited to Windows NT PCs;
The Software was limited to Oracle and MS-Access for data management, MOVE3 for Surveying data control and field book calculation, GCarto for cartographic editing and vector/raster management, Visual Basic and Visual C++ for development, ArcCAD for building topology, ArcView for querying.

CONCLUSION

The system, although based on entry-level hardware, is still running and more and more implementations have been carried out in the last period by local technical staff trained during the pilot phase. The decisions to move the registration under the Ministry of Justice has slowed down the process, which is a further proof that technology is far ahead politicians‘ designs.

FULL TOPOLOGY GIS AND RELATIONAL DBMS: THE CASE OF BULGARIA

ORGANISATION, FUNDING, HUMAN RESOURCES, DATA SOURCES
The European Union and the Government of Bulgaria in 1995 have recognized the need to help land reform and the establishment of a land market with a suitable IT infrastructure. Therefore, a PHARE project was started to provide funding and consultancy services for setting some experimental cadastral offices.
In the making of this computerized system, some issues were raised by the partitioning of tasks between two different organizations (Ministry of Justice and Ministry of regional Development and Public Works). Given this partition the system had to be built around two distinct data bases, which could, in the future, become two different instances of a single database (by means of replication).

GIS AND RELATIONAL DBMS: A PERFECT MARRIAGE
Managing geometric data, which is the core of GIS technology, was (and still, mostly is) done by giving an identification number (ID) to each geometric feature and linking every geometric feature to a record containing its alphanumeric data (attributes). Nothing can be more different from the Relational Data Model, which doesn’t use IDs and uses composition of attributes (known as keys) for identifying single records (which are, indeed, called tuples).
Cadastral applications comprise the managing of data which are both geometrical ad alphanumeric, moreover, a strong emphasis is put over reliability and integrity of data (especially legal data), hence four alternatives are given:

1. To use an entirely GIS-based data model
2. To use an entirely relational data model (aka Universal Database)
3. To use an Object-Oriented DBMS
4. To use an hybrid approach: GIS for geometric data coupled with a Relational DBMS for managing alphanumeric ones.

The last choice is the best of two worlds, but require good communications between the two halves of the system. Luckily, modern GIS software packages have very good links with Relational DBMS. In the case at hand a combination of ArcInfo GIS software and INFORMIX Relational DBMS was chosen. This couple get on well each other, all the INFORMIX functionality are available in ArcInfo and, given the use of Stored Procedures, it was fast and reliable too.

THREE PROJECTS AND ONE SYSTEM FOR ALBANIA

ORGANISATION, FUNDING, HUMAN RESOURCES, DATA SOURCES
A PHARE Land Use Policy (LUP) Project has been funded in 1995 to define guidelines in management of resources and decision making for Albanian agriculture. The first aim of PMU/LUP was to retrieve all the existing but heterogeneous data (soil, properties, infrastructures, meteorology, irrigation, elevation etc.) and to store them in a computerized system.
At the same time the privatization program made necessary the creation of a new immovable property registration system (IPRS) to provide owners with secure rights. IPRS PHARE program is assisting the review of surveying procedures, of base map, of mapping activities and quality control; while IPRS USAID inputs have been directed towards the Land Registration. The construction of a dynamic integrated Cadastral and Registration System is the final practical goal of PMU/IPRS.
Scarce possibilities of control and laissez-faire policy caused serious, irreversible damages to large forests. For these reasons, another EU PHARE project was financed for forestry, focused on woods census and protection.

THE SYSTEM ARCHITECTURE
By now the three PHARE projects- Land Use Policy, IPRS and Forestry are collaborating to the realization of an integrated GIS , physically installed in the offices of the Ministry of Agriculture.
The design establishes for the final system a client/server Windows NT/Windows ’95 architecture, based on ArcInfo and Oracle RDBMS. A LAN will connect the Server with the at least 4 client sites. One client site is the Mapping Production Unit where devices for data capture and data plotting will be placed. The three clients/users (the Immovable Property Registration System, the Soil Institute and the Forestry Project) will be equipped with ArcView to view/elaborate maps and to aggregate/disaggregate tabular information.

UNIVERSAL DATABASE ARCHITECTURE
The system being built is the first nucleus of one ambitious project which will use an universal (spatial and alphanumeric) database (candidates products are ORACLE and SDE), based on an integrated data server with the following main responsibilities:

1. retrieving and distributing geographic and alphanumeric information to clients in a spatial/temporal schema
2. granting to clients different levels of access for different sectors of data base: from simple alphanumeric consultation, to integrated data editing and update
3. managing locks and concurrence privileges: every access to information is negotiated depending to privileges and priorities
4. emancipating clients from any physical organization of data (i.e. sheets, administrative boundaries) using indexes for both alphanumeric and geographic data
5. allowing variable-scale spatial elaboration (from small scale for wide phenomena geographic analysis to large scale for administrative certifications)
6. emancipating clients from software dependencies using standard data connectivity protocols
7. allowing integrated spatial and attributes analysis
8. managing clients "proprietary partitions" of data, to study temporal evolution
9. mailing information in a suitable format to other Government agencies.

PRESENT SYSTEM DEVELOPMENT
During this year (1998) PMU concluded the phase of analysis of requirements (including technical specifications for maps digitizing), the phase of experimentation for cadastral maps digitizing and testing and finally started the phase of sub-contracting cadastral maps digitalization. Land Registration Project, with USAID supervision, is completing the onerous work of titles recording and controlling. A special software, named Manaxher , based on ArcView and on Microsoft Access, was developed by PMU to schedule the activities of subcontracting maps digitalization. Manaxher allows to control and to document source maps origin, age and quality, problems of matching on the borders of maps, problems of congruence between maps information and registration data, quality tests on digitized maps.

WHICH STANDARDS TO EMBRACE IN A CADASTRAL GIS

GOALS OF STANDARDS IN A CADASTRAL GIS
In spite of its theoretic and abstract formulation, standards adoption has been a very practical necessity for IT specialists involved in the described projects. The necessity of receiving, testing, storing information of all different kinds, quality, age and formats, stressed the necessity of establishing rigorous criteria for data organization and manipulation. The number of different people, with different languages, skills and habits involved in the projects imposed the use of standards to freeze user requirements, describe data and function model, test and accept data and software.

STANDARDS IN DATA MANIPULATION: TC/211 AND OGC
Data structuring was faced in the projects distinguishing two order of problems: the physical formats for data collection, and the adoption of standards for data management. At present, all systems are provided with tools to import/export data from different well known formats for both graphic and tabular information. In our work we mainly referred to two sources: the International Organization for Standards Technical Committee 211 (ISO TC/211) and OpenGIS Consortium (OGC). ISO TC/211 is much closer to data and metadata while OGC is more focused on object technology.
The TC/211 bases its activities on the definition of a reference model, profiles of users, terminology, conceptual geoinformation schema, spatial, temporal subschemas, rules for application schema, quality, positioning services, geodetic reference systems, encoding, spatial operators, services.
The OGC bases its activities on the definition of a general and common set of basic geographic data types, which can be used to build interfaces between dissimilar geoprocessing systems. The result will be a software schema in which data and processing functions are packaged into small, discrete, interoperable modules, offering advantages such as portability and easy maintainability.

STANDARDS IN GIS SOFTWARE
The last ten years saw the success of commercial GIS packages versus home-grown systems. Some commercial packages data formats became standards for information interchange; big companies opened branches for technical assistance in nearly every country. On the other side, software-engineering standards need big investments, generally not affordable by little companies, which, in fact, since the beginning of ‘90 started to focus on customization and assistance. So the choice of GIS software was restricted, in all these three experiences, to a certain number of high quality, wide diffused, packages.

STANDARDS IN SYSTEM DEVELOPMENT: ISO9000 AND IEEE
The activity of software development can be viewed as an industrial process and ISO 9000 guidelines may be applied to raise the quality of the final product (the system being built). For the ISO 9000 guidelines to be applied to system development, some specific best-practices guides should be considered. The main sources utilized to the development of application software in the three described projects have been the Institute of Electrical and Electronics Engineers guides and standards. These publications describe sound engineering principles in software engineering for every system development phase. IEEE standards indicate four phases for software development: Requirements, Design, Coding, Test. Requirements contain what the system has to do according to real needs and must be defined according to very strict rules in a document called Software Requirements Specifications.
Design phase output is described in (5) and must be carried out using widely-accepted methodology, (Entity-relationship diagrams, Data-Flow Diagrams, Unified Modeling Language). Coding phase is to be carried out using sound programming practices like modularity and readability in traditional development or encapsulation and overloading in modern systems. Another standard covers the testing phase (please, refer to (6)) and it is the right place in which the system is verified.

RESISTANCE TO STANDARDS ADOPTION

The adoption of standards encounters, the world over, obstacles and resistance; no wonder our experience in former-communist countries has been peppered with opposition, when it came down to standards.

The usual objections were:

1) This is not our way of doing things
2) Commercial software costs a lot more than hiring local programmers and making them build software from scratch
3) Our data standards are better than, say, DXF

Our response to these objections was given along following lines:

1) Of course this is not your way of doing things, neither is ours. Nevertheless, these guidelines for system development and quality assurance has been validated by international organizations and meet best practices produced by professional bodies; hence, they give a better chance to make a system work.
2) Yes, without any doubt, standard software (like ArcInfo) costs a small fortune, especially in poor countries. However, buying a standard GIS package means knowing all the cost in advance, while "reinventing the wheel" means, as home-grown software grows more complex, spending an increasing amount of money in maintenance.
3) Given that GIS applications in former communist countries are still in their infancy, it is much better to start with an early adoption of de-facto (Shapefiles, coverage, DXF) or institutional (VPF, Digest family) data standards. This may cost a little more now, but will allow reaping the benefits of using off-the-shelf tools and plenty of skilled workers later.

It is worth noting that, often, these objections come, understandably, from local software houses, but they strike a chord in politicians’ hearts, because they appeal to national pride and protection of local producers.

CONCLUSION: A CHANCE FOR EASTERN EUROPE GIS INDUSTRY

The convergence of East European countries to the European Union can be made easier by the convergence of their GIS industries. To a foreign investor a country with a good digital cartography is a country that is easier to invest in, say, agriculture or mining.
Of course, good cartography means one that is easy to use, hence, one that is data-standard-compliant. Local industry should see standard has an opportunity: given the low cost of skilled GIS professionals in these countries, this industry may grow remarkably in the future by reaping outsourcing contracts for data input and software development from western firms.
This growth may happen provided these conditions are met:

1) Confidence of western firms in former communist countries’ companies: this confidence can be bought adopting internationally recognized quality standards and following industry’s best practices
2) Skills that can be sold in western countries: hence focusing on GIS packages and development tools that are common in western countries

Therefore, this may be the right time for GIS industry in former communist countries to grow, provided it embraces standards and tools which are common worldwide, and leverages its valuable human resources.

REFERENCES

1. ISO 9000, International Standards Organization
2. ISO 9000-3, International Standards Organization
3. IEEE Recommended Practice for Software Requirements Specifications, Institute of Electrical and Electronics Engineers, Std 830-1983, IEEE 1983.
4. IEEE Standard for Software Verification an Validation Plans, Institute of Electrical and Electronics Engineers, Std 1012-1986, IEEE 1986.
5. IEEE Recommended Practice for Software Design Descriptions, Institute of Electrical and Electronics Engineers, Std 1016-1987, IEEE 1986.
6. IEEE Recommended Practice for Software Test Documentation, Institute of Electrical and Electronics Engineers, Std 829-1983, IEEE 1983.
7. LKI -Cadastral LIS The Netherlands- B. van Osch, Dutch Kad –ITC, 1996
8. Land Information in Eastern Europe, S. Hartley – GIM,1997
9. European Standard of GeoInformation (European Committee for Standardisation) Document prEN12160, CEN-Brussels,1995
10. The Digital Geographic Information Exchange Standard (DIGEST), DGIWG,1992
11. A GIS Approach to Land-Use Change Dynamics Detection, C.P. Lo, R. Shipman - PE&RS, 1990
12. Cost effective data for regional GIS bases, L.H. Spradley - ISPRS, 1994
13. Feature level topology for ArcInfo using the ArcInfo-Oracle interface, J. van Smaalen, M. Molenaar - ISPRS,1994
14. Design of a National Topographic database: an Object oriented approach,B. Shahrabi – ISPRS,1994
15. Framework report on database management systems - ANSI/X3/SPARC, 1978
16. Proceedings of the international workshop on object-oriented database systems, K.Dittrich - ANSI-AFIPS, 1978
17. The OpenGIS Guide,K. Buehler, L. McKee – OGIS, 1996
18. ISO/TC211 work programme - ISOTC211, 1995
19. Automation of the Cadastre in El Salvador, A. Tjalma - The Dutch Kadastre, 1995
20. The Thailand Land Titling Project P. Angus-Leppan INT. GEOG. INFOR.SYS, 1989
21. GIS versus CAD versus DBMS:What are the differences?, D.Cowen - PE&RS, 1994
22. A Perspective on GIS Technology in the Nineties, A.Frank,M.Egenhofer - PE&RS, 1991
23. The Indonesian Land Administration Project, C.Grant – GIM, 1997
24. First Generation OpenGIS Components, J. Roodzand – GIM, 1996
25. Towards Integrated Land Registration, G.Remetey – GIM, 1996
26. Views on Iteroperability and Spatial Data Infrastructures in Europe –GSDI and ESDI P.Burrough – ASITA, 1997
27. Quality in GIS, L.A. Koen – ASITA, 1997
28. Towards a European Policy Framework for Geographic Information DGXIIIe Luxembourg, 1997