Head, ESRI Consulting, ESRI India
The methodology of software development has matured a lot in the last two decades. A very large number of researchers have spent many man-years in developing the standards and procedures for successfully evaluating and implementing software projects. The GIS based projects have some commonality with the standard software projects but at the same time there are major differences. GIS based projects involve data creation including updation of data with remote sensing tools, data organisation, application development and website development. In this paper we have made an attempt to analyse various aspects associated with GIS based projects.
GIS based projects have major components i.e., data development, data organisation and application development. In some cases the client needs a website developed also for better proliferation. A large number of organisations using GIS softwares are in public sectors while some are in private sectors. Many of the organisations have a lot of software knowledge while some may not be that accustomed to software development life cycles. Any knowledge based be it software or GIS needs a very clear interaction between the developer and the client. There has to be a very strong need for discipline, mutual respect and will to succeed from both sides to execute a project successfully. If a project is assessed properly and boundaries well understood by both the client and the developer then the projects can be successful, if not it can be disastrous.
Development of GIS
Geographic Information Systems (GIS) are computerised systems for the storage, retrieval, manipulation, analysis, and display of geographically referenced data. GIS includes physical, biological, cultural, demographic and economic information and hence they are valuable tools in the natural, social, medical, and engineering sciences, as well as in business and planning. The fields of cartography and mapping have been practiced since a long time but GIS as such is a comparatively new field.
Researchers have identified three research areas as components of a critical social history of GIS. The first involves precursors and preconditions for the development of GIS. These include the intellectual and technological prehistory of GIS including pre-modern and early forms of GIS. These issues require an in depth understanding of history of how systems of representing geographic information developed, such as metrication, land surveys, military surveillance, and mapping expeditions. A second area of importance in the history of GIS is the history of applications of GIS in different cultural and political economic contexts. The third key area is the development of contemporary GIS software and data. Although some new GISs have been developed starting with a clean slate, current GIS is dominated by software and data models whose intellectual and conceptual lineage can be traced directly back to innovations in the 1960s. The nature of current GIS cannot be fully understood without a historical perspective on how it got to be, the way it is.
Development of Geographic Information Systems started in the early sixties and was initially encouraged by the public sector. However, in the 1970s and 1980s, a vigorous GIS industry developed, in both private and public sectors. GIS software, data, and services is known to be a $2 billion industry at present, and the industry is believed to be growing at about 20 per cent each year. GIS also has had a major influence on the discipline of Geography in the 1980s and 1990s, variously seen as a critical factor in reviving academic geography.
Acquiring data for use in GIS can be the most costly and important part of GIS application development. A GIS, which can use information from many different sources, in many different forms can help with analysis like which watersheds should be prioritised, which land can be used for cultivation etc. The primary requirement for the source data is that the locations for the variables are known. Location may be annotated by x, y, and z coordinates of longitude, latitude, and elevation. Any variable that can be located spatially can be fed into a GIS. Many government agencies and private firms produce several attribute and spatial databases that can be directly entered into a GIS. A GIS can also convert existing digital information, which may not yet be in map form, into forms it can recognise and use. For example, digital satellite images can be analysed to produce a map like layer of digital information. Likewise, tabular data can be converted to map-like form, serving as layers of thematic information in a GIS. New data can be created using satellite images, field survey data and GPS survey data.
There are different sources of data that can be used for creation of digital data. Some of them are listed below:
- SOI Map – Different scale maps
- Updation using satellite imagery
- Maps from other sources (AISLUS, GSI, FSI, NATMO)
- Cadastral Map
- Data creation using satellite imagery
- Data from GPS
Important Issues for creation of data
- Need Analysis: The client will have certain ideas for the data development. However, in most of the cases he may not be fully clear or aware of what best can be done. It is the job of the GIS analyst to carefully assess the requirements of the user and suggest what is best suited for his needs. Such an analysis will help in evolving the needed specifications.
- Cost: With any type of information system, careful planning prior to data acquisition or data creation generally increases the degree of success and utility. Lack of planning increases the possibility of a mismatch between a system’s capabilities and user needs, resulting in wasted money. The cost of data development is quite high and the customer must be sensitised and shown a method by which he can achieve his aim within his estimate.
- Accuracy: Another major issue for data creation relates to accuracy, completeness and timeliness of the. Digitising information for inclusion in a GIS is not simple and straightforward. As information is included or excluded to fit with the application, the accuracy and completeness of that information may be compromised. In addition the conversion of existing records, without verifying the accuracy of the information with the data subject, may mean that the quality of the newly digitised information is poor as it is out-of-date or incomplete.
- Scaleability Issues: It is important to develop data in such a manner that future expansion is possible. If this is not considered during data creation then different kinds of spatial data pertaining to the same area cannot be attached with each other due to difference in projection systems and attribute information.
- Lack of quality and availability of base data: Data quality is essential when a GIS is used to make decisions that, potentially, could adversely impact the data subject. Without accurate information, any potential operational efficiency or benefits, may be compromised. Lots of data creation effort gets stuck because either the base data being unavailable or of very poor quality.
- Permissions from regulating authorities: This policy on data development needs a relook.
- Standards/Formats: A standard format needs to be developed for spatial data as data conversion from different formats causes a lot of data loss and the quality of the data gets reduced.
- Symbology: Standard sets of symbols should be developed to use with all kinds of applications using GIS.
Structuring of the database and their indexing and other related issues play an important role in the GIS application development lifecycle. Just as in any normal database activity, the GIS database also needs to be designed so as to cater to the needs of the application that proposes to utilise it. Apart from this the design would also:
- Provide a comprehensive framework of the database.
- Allow the database to be viewed in its entirety so that interaction and linkages between elements can be defined and evaluated.
- Permit identification of potential bottlenecks and problem areas so that design alternatives can be considered.
- Identify the essential and correct data and filter out irrelevant data
- Define updation procedures so that newer data can be incorporated in future.
The design of the GIS database will include three major elements
- Conceptual design, basically laying down the application requirements and specifying the end- utilisation of the database. The conceptual design is independent of hardware and software and could be a wish-list of utilisation goals.
- Logical design, which is the specification of the database vis-a-vis a particular GIS package. This design sets out the logical structure of the database elements determined by the GIS package.
- Physical design, which pertains to the hardware and software characteristics and requires consideration of file structure, memory and disk space, access and speed etc.
Data stored in an inefficient framework would cause performance overheads. The success or failure of a GIS project is determined by the strength of the design and a good deal of time must be allocated to the design activity.
GIS application development is basically customisation of existing GIS software to meet specific needs. The applications may be as simple as a set of preferences that are stored for each user group or individual. Or they may be a very complex queries that selects a group of layers, identifies features of interest based on attribute ranges, creates variable width buffers, performs a series of overlays and produces a hard copy map. In either case, an application is required to convert the user’s ideas into a usable, stable product. In a large number of cases after using simple application, the users become more experienced with GIS and then they require more complex applications. The initial needs assessment will contain some applications of a complex nature, however the majority of initial applications will be straightforward, using the basic functionality that is part of every commercial GIS (e.g., query and display). The more complex applications usually are not supported by the basic functions of a GIS but must be programmed using the GIS macro language or other programming language. This section identifies several categories of applications that can be prepared by users and how overall requirements change over time.
Applications are usually built on existing GIS packages. These packages provide the functionality common to all user disciplines. Commercial GIS packages tend to focus on the common or basic applications. When it comes to specialised uses, application development fills the needs for functionality. Though there is a great deal of commonalty in the basic spatial queries and display functions, there is still a need for other advanced applications. We need additional applications because needs are different between organisations.
Types of Applications
Product Development: There are many GIS products available in the market today which help users to perform spatial queries and analysis. The first of such products was a grid-based mapping program called Synagraphic Mapping Sytem (SYMAP), which was developed at the Laboratory for Computer Graphics and Spatial Analysis at the Harvard Graduate School of Design in 1966. This system was widely distributed and served as a model for later systems. These early GIS packages were often written for specific applications and required the mainframe computing systems found usually in government or university settings. In the early seventies private vendors like Environmental Systems Research Institute (ESRI) and Intergraph began offering off-the-shelf GIS packages. Over time ESRI has emerged as one of the leading vendors of GIS software. In 1981, ESRI released Arc/Info, a standard package, which ran on mainframe computers. As computing power increased and hardware prices plummeted GIS became a viable technology for all kinds of organisations. In 1992, ESRI released ArcView, a desktop mapping system with a graphical user interface that marked a major improvement in usability over Arc/Info’s command-line interface. The development of ArcView for Microsoft Windows and ArcIMS, which enables distributed mapping and spatial analysis over the Internet and eliminates many of the hardware and licensing expenses of a full software package, has increased the availability of spatial data to marginalised and under funded groups. Other GIS softwares are also used by various users.
Information systems: Information systems are GIS applications, which can retrieve information about a particular area, location or object. Information systems are used for land use planning, utilities management, ecosystems modelling, landscape assessment and planning, transportation and infrastructure planning, market analysis, visual impact analysis, facilities management, tax assessment, real estate analysis and many other areas. The functionality of such systems is varied. It can be as simple as finding the location of a particular place that meet a specific criteria or where the most number of a particular kind of features lie. For example, if a person wants to find out all the hospitals within a 2-Km radius of an accident area or the number of hotels nearest to the railway station then he can use a system like this.
Decision Support Systems: Another way to classify a GIS application is by the kinds of decisions that can be taken using it. Individual GIS systems are generally used to make several different kinds of decisions like what kind of proactive actions need to be taken in case of a natural disaster or what policies should be developed for the improvement of natural resources etc. GIS decision support systems are created using different kinds of algorithms, equations and analysis model. For example the Djikistra algorithm for finding the shortest route, or the Universal Soil Loss Equation (USLE) to calculate the probable soil loss for an area due to erosion etc. Such systems also have wide applications in the field of defense, emergency services, healthcare, conservation of forests etc. Lets take an example of GIS decision support system used for business. Such a system can be used by a manager to answer questions like, where are my customers located? What are their characteristics (market segmentation, classification of residential areas)? Where are my competitors located? What is the potential turnover in a region for my product? What market share can I expect? Where should I locate my new branch? Should I expand an existing branch? The answers to these questions will then help him to take many tactical and strategic decisions.
Decisions can also be made using modeling tools. Modeling tools are one of the most powerful and sophisticated tools available to GIS. There are three types of modeling: cartographic, simulation, and predictive. Cartographic modeling is more of a project organisation technique and hence it will not be discussed here. Simulation modeling involves using the GIS to simulate a complex phenomenon in nature. This generally requires an expert to create, and can vary in the degree to which it is linked to the GIS. However, once the GIS and the model are linked, they can be used to evaluate different features of the data, whether it is spatial or non-spatial. The other, more powerful modeling tool is predictive modelling. In this form of modeling, an expert acquires data and uses it to build a statistical model, which is tested by regression analysis. Once the model has been tested on known data, it is applied to new data in order to predict results. This type of modeling has been used to predict flooding, groundwater contamination, and soil loss, to name just a few. The ability to link GIS to these models has greatly increased the usefulness of GIS as a scientific decision making tool.
Software for Data Creation and Development: Many utilities are created using the macro languages of existing GIS packages or any other programming language to assist in data creation and development.
Issues for Application Development
System Requirement Study (SRS): An SRS is the first step in implementing a successful GIS within any organisation. An SRS is created to assess the needs of the user. It is a systematic look at how departments’ function and the kind of spatial data needed to do their work. At the conclusion of a needs assessment, an organisation will have all of the information needed to plan the development of a GIS system. The SRS functions are same as in the case of software development however there are some specific issues pertaining to GIS. This information can be grouped into the following categories:
- Applications to be developed: In evaluating the responsibilities and workflow within a department, certain tasks are identified that can be done more efficiently or effectively in a GIS. These tasks will form the basis of GIS applications. Application descriptions prepared as part of the needs assessment will describe these tasks.
- GIS Functions required: For each application identified, certain GIS functions will be required. These will include standard operations such as query and display, spatial analysis functions such as routing, overlay analysis, buffering, and possibly advanced analysis requiring special programming.
- Data needed in the GIS database: Most organisations going for a GIS system, use data that has a spatial component. Much of this data are hard copy maps or tabular data sets that have a spatial identifier such as addresses and pin codes or X-Y values (latitude-longitude, state plane coordinates, or other coordinate system). A needs assessment will identify how this information will be used by GIS applications.
- Data maintenance procedures: By looking at the workflow and processes within and between departments, responsibility for data creation, updates and maintenance will become apparent.
All these aspects have to be covered in the SRS phase itself for the GIS application to be developed pragmatically and with minimum defects.
Tender Issues after pilot project creation- In some GIS projects organisation ask a vender to develop a prototype but when the time comes to actual development are forced by their procedures to get into tenders. This causes serious problems for prototype developer, it is essential that his interests are protected.
Freezing of GUI Designs- Success of the project depends on early freezing of GUIs, as it gives confidence to the vender and avoids rework of the vender.
Website development- Many of the GIS projects need a website development. It is important that this need is finalized in the beginning and entire development is carried out keeping this in mind. We have seen successful projects failing, as the websites were too slow. One of the reason is lack of optimisation in database development and coding while creating the application. Other issues that require to be considered are the hardware configurations and bandwidth related issues.
Time and cost- Large projects must be taken up phase wise. This gives satisfaction to vender as well as to vender due to results achieved and cash flow accrued.
Hardware Issues- Most of the commercial GIS packages and customised applications have prerequisite hardware requirements that need to be satisfied for them to function properly. Before the initiation of project development a decision has to be taken as to what kind of hardware (minimum) has to exist for the application to work. For example for a simple mapping application a system with Pentium processor, 32 MB RAM and 2 GB hard disk will suffice but an application having complex queries and analysis requires at least a system with Pentium III processor, 256 MB RAM and 10 GB of hard disk space. These issues need to be resolved at the inception stage itself or they cause problems during the implementation of the application.
This paper by no means gives a comprehensive treatise on GIS projects management, it is only a pointer to some of the issues which if taken care of at an appropriate time will help handle GIS projects better.