It is very commonly stated that "Drawing is the Universal Language of Engineering". All engineers understand the importance of drawing and the representation of graphical information for engineering decision making. With the advent of the digital age, drafting tools such as Autocad provided platforms for digital representations of drawings. In parallel, databases software such as Oracle provided platforms for representing text and numerical (nongraphical) data such as payroll etc. Although, engineering decisions required analysis of both graphical and non-graphical data; in the early years of the digital age, no single software platform was available to integrate graphical & non-graphical data. Thus non-graphical data had to be noted on the drawings or other documents and human intervention was essential to integrate the information.

In its most basic form a Geographic Information System provides an integrated software platform for representing graphical and non-graphical data. In addition, this integrated data representation can be queried based on graphical and non-graphical attributes.Another essential feature of most GIS software is the ability to perform spatial analysis such as buffering and overlay which enhances the decision support capability of the tool. More sophisticated GIS software provides advanced features such as routing, surface modeling and customization.

Corridor selection or routing is an important decision on many infrastructure projects. Crosscountry facilities such as roads, railway lines, pipelines,canals, transmission lines need to be routed appropriately. The route corridor can have a significant impact on the cost-effectiveness of the projects as well as the success of its operations. This article discuses a structured GIS based methodology which can be used for corridor routing and illustrates how the methodology can be used for routing transmission lines and pipelines.

It should be highlighted that while the methodology and features of the software used can be used for routing, the quality of the corridor depends on: (i)ability of the decision maker and GIS analyst to identify, quantify and represent all the factors influencing the route and (ii) accuracy of the data represented in the software.

General Methodology

A generic raster GIS based methodology for corridor selection any infrastructure feature is shown in Figure 1. In the first step, the specific factors to beused for routing have to be specified by the planners and engineers. In the second step, data pertaining to these factors has to be gathered. In the third step the data is entered into the Geographic Information System. At the end of this step, the integrated database will be available for querying and analytical operations. For routing, additional data on the priority of the factors for route analysis is required.This can be represented in the form of weights.

Determining the weights for each factor can be a challenging task. This essentially requires experienced planners to translate their tacit knowledge into an explicit formulation through quantification. It is common to assign these weights through intuitive means. However,structured approaches such as Analytical Hierarchical Process (AHP) [1] has shown to result in more rational assignment of weights for various factors.

Once the weights for each factor are determined,it is integrated along with GIS analysis operators such as buffering and overlays to develop a suitability raster map. This raster map will divide the study area into specified number of grids and each grid will be assigned a weighted cost based on the features it contains and the weights assigned to the features. For example if a grid is in an area with steep slopes and steep slopes are to be avoided it will have a high cost.A grid covering a forest area with steep slopes will be associated with an even higher cost as it has two features which are not desirable.

The cost-distance map utilizes the Suitability Map and the specified origin/destination as a inputs and assigns each grid cell a value of the minimum cost to travel from the grid cell to the destination. By default this cost can be based on Euclidean distance.But as the cost of each grid cell is established in the suitability layer, it is more practicable to base it on this weighted cost.

The cost-direction map is an extension of the costdistance map and each grid in this raster map will be associated with a single direction of travel to reach the destination from that grid cell at minimum cost.Once this map is generated, the minimum cost path to the destination can be determined by linking the series of adjacent grid cells between the origin and destination as specified by the direction in each grid cell.

This general methodology is illustrated in the next section using two examples. Both studies are a part of projects done at IIT Madras guided by the Author. The first is a hypothetical example of a pipe corridor routing; more details of this work are given in [1].The second study second is a case-study of a transmission line routing. More details of this study are given in [2]. These studies were done with support of engineering managers from Larsen & Toubro  Construction Group. Mr. K.S. Suresh supported the pipeline study and Dr. K. Natarajan supported the Transmission Line Study.

Pipeline Corridor Selection

The very first step in pipeline corridor selection was is to identify the global location of the source & destination. The location of the source and destination aided in scoping the study area. Within the study area selected, the following factors were considered to influence pipeline corridor selection.

•  Avoid areas with steep slopes
•  Avoid road crossings
•  Avoid railway crossings
•  Avoid river crossings
•  Avoid prohibited areas
•  Avoid reserved forest areas
•  Avoid areas with high land cost
•  Avoid unfavorable soil type

To include the above factors it is apparent that geographical data relating to the contours, roads, railways, rivers, and other factors need to be collected. Some of the data can be collected from toposheets and satellite images, but field data collection is also essential for validating the data and collecting data such as soil type which may not be available from secondary sources. Once the data is collected it has to be organized into appropriate GIS layers and suitably digitized and integrated with required nongraphical attributes to form the GIS database.

Figure 2 shows a suitability map for the example problem. The Cost-Distance function and the Cost Direction function are successively applied with the source as the target (indicated with a green plus). The resulting cost-distance map and cost-direction map are shown in Figures 3 and Figures 4 respectively. 

The cost-distance raster gives the accumulated cost of each cell on the suitability map from the origin. The cost-direction raster provides the road map, identifying the route to take from any cell, along the least cost to the origin.

Transmission Line Corridor Routing: The factors considered for selection of transmission line corridor are shown in Figure 6. The figure also highlights the spatial operations performed on each of the factors. As discussed earlier, weights were assigned to each of these layers based on transmission line routing requirements and the overlays were performed to generate the suitability map as shown in Figure 7.
Only a limited corridor within the area is taken for analysis as the remaining areas were constrained due to other factors which were identified in the initial portion of the study. The origin point and destination point shown in the figure represent the start and end points of the transmission line. Figure 8 and Figure 9 shows the cost-distance map and the cost-direction map for this case. The final route of the transmission line corridor overlaid on a  satellite image of the area is shown in Figure 10. Further details of their study are available in [2].

Summary

This article highlighted and how Geographic information Systems can be used for optimal corridor routing of cross-country infrastructure project. A generic methodology based on raster GIS was presented to implement the process. The application of this methodology was illustrated using pipelines and transmission lines as an example. The same methodology can be applied to route any cross-country infrastructure facility such as roads, canals,. While the GIS technology provides a powerful platform for the analysis, identification of the factors influencing the path of the facility, gathering accurate data and establishing the weights for the factors are the critical steps. If these steps do not reflect the site situation realistically, the routing suggested will not be optimal to meet the project requirements.

In addition to routing applications there are a number of other applications in construction which a GIS can be used for. Some of these applications are highlighted in [3], [4] & [5]. GIS provides a powerful platform for spatial querying, visualization, analysis and decision support, the construction industry should make a focused effort to incorporate this powerful technology into the project decision support processes.

• Colin Nonis, " Investigation of an AHP based Multicriteria Weighting Scheme for GIS Routing of Cross Country Pipeline Projects" M.Tech Thesis, Department of Civil Engineering I.I.T. Madras 2006
• Syed Abdul Rehman "Study of GIS Application in Transmission Line Siting and Optimization" M.Tech Thesis, Department of Civil Engineering I.I.T. Madras 2004
• Manoj, P., Mahadevan, N. & Varghese, K. "Optimal Layout of facilities on a Construction Site: A Genetic Algorithm Approach?,ASCE Computing Congress, Philadelphia, 1997
• Varghese, K. "Decision Support On Construction Projects Using Geographic Information Systems" 1st International Conference on Construction Project Management?Jan 11-12, 1995, Singapore.
• Shanmugam, S. P., Parvatham, R. & Varghese, K. "Geographic Information System for Coordination of Fast-Track Projects", Journal of Computer-Aided Civil and Infrastructure Engineering, Vol. 17, No. 4, 2002, 294-306.