Geospatial technologies integration / management

Geospatial technologies integration / management

Ricardo Salazar
P.E., GIS Project Manager
Metropolitan Water District of Southern California
700 North Alameda Street
Los Angeles, California 90012

Introduction
The primary concern for a Project Manager when managing a GM project is to utilize the technology to its full extent. Several concurrent technologies can be used and combining them will impact all project phases, moreover they will impact the usability and life of the application, total cost and return on the investment.

Being responsive to client needs and any organizational changes required to meet those needs should be a priority in today's highly competitive environment. Applications, systems and data integration that will allow expedient and accurate access to the information is a vital component of application development.

The evolution and availability of non-traditional types of data such as CAD drawings, raster images, video, spatial information or in general complex data, has generated end user demand to use these sources to extract better information. For example, a potential customer looking for services in the web or an employee searching internally for information, prefers to use an application with a graphic interface for ease of operation. By not providing this capability there is a high risk of losing the customer. New technology that will allow us to manipulate complex types of data as objects, needs to be used as we see in the new concept of Universal Servers that is an evolution of the relational databases. Knowing when to use them and the available options is very important for the design of the applications.

Two other technologies related to GIS that we need to be aware of, are aerial mapping and remote sensing. Proven benefits must be weighted, and a decision made to use either or both technologies. The right combination will provide us a better and cost effective solution. Current in-orbit satellites fiu-nish a variety of spectral and spatial data, depending on the application the use of one or more sources may be more beneficial. Considering that normally the information obtained directly from the satellite companies is raw data or "level 1", we should also take into consideration independent companies that acquire data from different sources and then process that data (Orthorectification, Data fusion Co-Registration, large scale mosaics, elevation extraction (DEM), tonal balancing) and deliver it in GIS ready format. The main variables to consider when defining the data to be used, are coverage, pixel size, resolution, availability, cost and accuracy. In a later portion of the paper we will discuss this issues and the impact that they have on the project.

We will also discuss document management and how this technology has been integrated with GIS. Often we see this technology imbedded in a GIS application. The tendency is to become a paperless society due to the advantages gained such as getting fast access to crucial information, reducing the cost of "handling" information, greatly improving information decision-making, decreasing the time workers spend processing documents and other factors. To make possible the integration of GIS and Document Management three component technologies need to be in place; 1. document management software, 2. imaging retrieval and manipulation software and 3. workflow software.

Another technology used by GIS is the Global Positioning System (GPS) that can be used to acquire different levels of information, from very accurate centimeter location to 2-5 meter accuracy. This capability will allow us to capture precise base map information using differential kinematic GPS, or city street lights mapping to one meter accuracy by using different GPS receivers and stationing times. A sample of dynamic integration of GIS and GPS will be a navigational system used by some rent-a-car companies to travel from one location to another, they show the city map and the actual location of the car, including instructions to get to the destination.

We also need to keep up to date with a variety of communication technology such as local, remote, mobile, and Internet technology, alone with an understanding of their impact in the GIS application. As part of the application design, we need to define where the data will reside, what type of data can be transmitted based on the availability, communications speed and cost. For example, to keep track of vehicle location containing hazardous materials, the communications aspect is the most costly, specially if it is in a remote area or one not covered with a telecommunications network. The Internet and object oriented technology has allowed us to deploy applications for internal and external users. Considerations of confidentiality, security and performance should be addressed.

For software systems that are very closely related with GIS like Computer-Aided Drafting (CAD), Facilities Management (FM), Supervisory Control and Data Acquisition (SCADA), we will see that their primary integration with GIS is as a graphic presentation or analysis tool for integration of graphic and text data. An important part of this discussion, is what needs to be considered in each system to facilitate the integration of the parts, to achieve an integrated information management system.

The discussion then will cover critical factors that should be considered within each technology, when evaluating what to use to satisfy the application requirements. By considering these factors, it will help us to achieve client satisfaction, integration with other company applications, low cost and a good return on the investment.

Most of the observations included in this paper are based in the present development of GIS at the Metropolitan Water District of Southern California (MWD). MWD provides approximately 60% of the water used by the nearly 16 million people living on the coastal plain of Southern California between Ventura County and the Mexican border, covering about 5,200 square miles and including 240 cities and unincorporated areas. Several reservoirs, filtration plants and an almost 1,000 mile pipeline distribution system that includes the Colorado River Aqueduct are used to satisfy the yearly average of 1.8 million acre-foot water demand (one acre foot= 325,851 gallons).

Technologies integration
The technologies that will be discussed emphasize their use for better integration and cost effectiveness when used to develop GIS applications. These technologies can be grouped in the areas of data capture (photogrammetry, remote sensing, GPS, CAD, SCADA, and EDMS), database design and implementation, and communications (local, remote, mobile, Internet),

Data Capture Technologies
1. Aerial Photogrammetry: From the different classes of photogrammetry we are interested in aerial photogrammetry, used to develop pkmimetric detail and topographic configurations. Several steps are involved in the generation of digital maps that we need to be aware of such as acquisition of aerial photography, field control surveys, digital data collection, aerotriangulation, photographic reproduction, orthophoto and feature capture field survey.
  
  The project manager has to decide how to use the technology after evaluating the requirements of the application. The main points to consider are the positional accuracy, the boundary for that accuracy, the pixel resolution and deliverables needed.
  The combination of these requirements will impact directly the cost of the project.
  
  For example, if we are dealing with a pipeline project, we don't need to provide the same accuracy to all the project. A greater level of accuracy can be provided for areas closer to the pipeline and lesser accuracy to the rest. The flight altitude will be the primary determinant of the accuracy, and will determine the ground area covered by each photo the number of photo / stereomodels and the density of ground control, issues that will impact the cost directly. Then the determination of the project boundaries with different levels of accuracy is very important.
  
  Another issue is the pixel resolution. Use the resolution that you need depending on the features that you need to identify. Don't go beyond that. Also consider other alternatives to capture some of the features that will help you going to a lower resolution. The pixel size will impact directly the size of the files where size= (length x resolution) x (width x resolution) x 1 byte/dot for an 8-bit gray-shade .TIF file. And for color the size will be triple. Now using algorithms we can compact the file sizes but still it is and issue that we need to deal with when we manipulate those files, specially if they need to be used in a network environment. Some schemes such as an image pyramid where different levels of resolution can be used to see an area depending on the level of zooming can be used. The farther you are when covering a big area the biggest size of pixels you will use diminishing the file size and manipulation time.
2. Remote Sensing: Multiple applications can be developed using this technology. GIS Environmental mapping, crop monitoring, Digital terrain modeling (DTM), and emergency response are a few samples of its use. For Engineering applications and mapping of large areas where aerial photography could be expensive or not available, remote sensing is a very good alternative.
  
  The generation of contour line maps from the DTM's, that in the case of the Systeme Pour l'Observation de la Terre (SPOT) imagery can give us a vertical accuracy of 7 to 12 meters, is used to detect the best alternatives to lay a pipeline or construct a road. These contours are also used for siting and system design of cellular communications systems, and for prevention and evaluation of floods and inundation caused by tsunamis.
  
  Factors that we need to consider when using this technology in addition to accuracy and resolution previously discussed are:
  1. The acquisition of the appropriate imagery that can be from multispectral scans, thermal scans, radar imagery or a combination from different sources of satellites such as LANDSAT, the IRS-1 A and IRS-2A from the National Remote Sensing Agency of India, SPOT, the National Oceanic and Atmospheric Agency (NOAA) and 2 meter resolution panchromatic SPIN-2 Russian satellite image data.
  2. Availability of imagery will depend on the coverage, revisit, sensor types, library available and age of the images.
  The combination of types of data is also important, for example for an emergency response application for a country we can use LANSAT data to cover general areas and we can use SPOT data for the city areas were the bigger losses will be. The usage of older data is also important, it can be used to generate a DTM because the topography doesn't change that much and four year old data will be half the price of newer data.
3. Global Positioning Systems (GPS): The usage of this technology as I mentioned in the introduction is very valuable for GIS applications. It can be used as a surveying tool, providing very accurate positioning and measures to establish the base map for engineering type of applications or to collect GIS attribute data for the features. This capability is the more important integration issue between GIS and GPS technologies, because we can efficiently collect and simultaneously position attribute data with the GPS data collectors.
  
  GPS has changed the way topographic and planimetric data is collected. The integration of GPS with other data capture technologies, such as laser and cameras, allows us to perform airborne laser mapping integrated with differential kinematic GPS with much lesser ground control. This has an advantage over previous methodologies, when collecting data for difficult access or remote located GIS projects.
  
  In both cases when we used GPS to collect accurate surveying data for our base map or attribute data for the features in the GIS applications, the main concern is planning which information needs to be collected. For example if we are collecting surveying information for our base maps to get a better location of our pipelines, we don't need to collect all the points of the pipeline. A combination of office and field work will give the best results in time and money. The key element is to select in advance which are the points that need to be collected in the field. This planning can save time and thousands of dollars.
4. Computer-Aided Drafting (CAD): Is used for editing and presentation of graphic information on computer-drawn maps or drawings. I am including CAD as part of the data capture technologies from the point of view of GIS because it is the main tool that allows us to enter vector data into the system. To better understand the integration of CAD with GIS, I will clarifi the use of the spatial data elements by each technology which are: Value(Database), Shape(Graphics), Location(Coordinates, Map Projection and Datum), Topology (Connectivity, Orientation, Adjacency, Containment).
  
  The spatial elements that CAD mainly uses are Graphics (Points, Lines, Areas, Symbols and Text), Coordinates and orientation. Topology is not used by CAD and Map projections are not of primary importance. Those two spatial elements are used by GIS. Then, if we consider that CAD will be feeding GIS with graphic and vector data we just need to see how this data is used by GIS to allow an easy integration between technologies.
  
  Because GIS uses topology and the main data elements used by GIS are (points, lines and polygons), one important task where CAD will ease the integration, is providing clean data needed by GIS to generate topology and allow spatial analysis, queries, routing, etc. Generation of CAD data elements must include:
  1. Closed polygons using precision placement where the last coordinate pair of a digitized area boundary must exactly match the first pair and all area boundaries that represent different features should reside on different levels whenever possible.
  2. Common line segments shared between graphical features should have only one line that represents the common segment.
  3. Linear features should be created using a continuous line string and when they intersect with other linear feature of the same type, level and symbology must be terminated at the point of intersection.
  4. Provide a list of the GIS items that will be used in the map and the levels to reside.
  Another important issue is the location of maps by using a predetermined coordinate system. There are more issues that need to be considered to assure a smooth integration of CAD and GIS to increase productivity and save time and money. My recommendation will be to implement inside the normal CAD standards used by the drafters the minimum additional standards needed to allow us to have CAD/GIS ready digital maps.
  
  By using these standards, you can alleviate the conversion process. In a study conducted in MWD implementing the GIS standards for the right of way maps mean an additional work of three hours per map versus ten to thirty four hours of conversion time per map needed in the future.
5. Supervisory Control and Data Acquisition (SCADA): SCADA can provide to GIS with data on current system status and also it is a source of historical data of systems operations. It is important to clarify that the type of information collected by SCADA is sequential in time for each point in the network, and it generates large amounts of information. In the case of GIS, we represent the information graphically for all the network at a certain point in time. Then the information that will be extracted from SCADA to be used in GIS will be more of an historic type, for example what were the pressure peaks in the network yesterday, or two hours ago. An online system is feasible, but more difficult to implement. By other side, GIS gives SCADA the graphic and spatial component and ability to perform spatial queries to the Operations area users. For example, if there is a problem reported by the public, that normally will provide street intersections or addresses to identify the location, then Operations by using GIS will be able to locate easily which segment of the pipeline corresponds to that intersection area and take the corresponding response measure.
6. Electronic Document Management System (EDMS): The need to access documents in a variety of formats from GIS is increasing because 70�/0 of all information used today has a geographic element and GIS provides the environment to do it. By integrating both technologies, a GIS user will be able to perform queries or access documents related to a geographic location which can be slides, digitized video and audio, scanned (raster) drawings and paper documents, and other digital documents.
  
  To allow the integration of GIS and EDMS our first task is to identify which are the documents that we are interested in to retrieving using the geographic component, how we will provide this identification to the documents, and how this geographic identifier will be included in the database. This identifier will be use when performing geographic queries, such as to retrieve documents contained in an area. An alternative can be to have a pseudo geographic search by linking our documents in the database to other features that cover our entire area to use their geographic component.
  
  The next step after the definition of the geographic identifier is to create the interface to communicate between the GIS and EDMS because EDMS normally has their own proprietary algorithms to save and access data. Also, when we came from the EDMS side and a user needs information with a geographic component like what documents can be related to a street intersection, we will need to send the query to GIS.

Database design and implementation
The majority of GIS applications were using proprietary Relational Database Management Systems (RDBMS) and as we know 70'XOto 80% of the cost when implementing a GIS goes to collecting cleaning and storing the data. The maintenance of this main asset needs to be through a technology that can manage different formats of data reliably, and with a low aging ratio to avoid maintenance expenses in upgrades and application changes. Databases of leading RDBMS vendors now are providing extended data support to store complex data types such as images, videos, CAD drawings and spatial information together with the corporate character data. They are known as object-relational databases or universal servers.

At the same time GIS vendors are providing spatial middleware solutions to use RDBMS to store their spatial data and allow traditional corporate users to perform geographic queries. In this moment we can say that both are partial and incipient solutions. The one coming from the RDBMS side and the other from the GIS vendor. For the short term, and if the company has already a big investment in GIS applications, the solution is the one that offers the GIS vendor by going through the middleware. In the long term the solution will be the one offered by the database vendors to manage the data. The GIS software will mainly be used to perform spatial analysis and GIS functions.

One interesting suggestion is to save the graphics information in tabular format. For example pipeline alignment information can be saved in tabular format and the graphics can be generated from there. Actual updates can be done directly in the database and any map using that alignment can be regenerated from there. As an additional advantage, this tabular information can be more easily integrated with corporate users using traditional tabular type of tools.

Communications
GIS applications can be local, remote, mobile, or a combination of them. Transmission capacity has been improving. Optical fiber's bandwidth supports data rates in the gigabit range. Multimode optical fibers can support transmission rates to 100M bps over distances of one mile without a need for repeaters. In GIS applications considering that graphics and raster data such as photos are involved, meaning large transmission data needs, it brings to our attention the fact that the speed of communications is a major concern for remote and mobile applications.

For local applications, where the servers are in the same building, and where general transmission rates for the end users are in the range of 10M bps the response time is acceptable. For remote applications normally connected using T1 lines with 64K bits of capacity, difference in transmission rates is substantial. The design of the application has to take into consideration this limitation and save or distribute those graphic files or themes more used in the remote locations to the servers located therein order to minimize transmission times.

If we need to enable mobile workforces to access important and often confidential information, in addition to the transmission rate concern for wireless applications we need to incorporate security capabilities in the design. This wireless solution will normally reside in the corporate intranet with efficient data compression algorithms and the data streams secured behind firewalls. The transmission rates in this case goes down to 33.6 kbps or 19.2 kbps.

Actually at the current speeds, e-mail or messaging type of applications works ok. For graphics applications this solution will be very slow. One possible way to overcome this difficulty will be by having the graphics residing in the laptop, and receiving only the tabular data and minimal graphics information from headquarters.

Conclusion
To be able to manage these different technologies that are used by GIS, we need to have an understanding of what are the advantages, and the technical and economical limitations of their use. Each technology has a set of factors or variables that will impact their use when used in conjunction with GIS. These concepts that I have emphasized along the paper will help us in making a better decision in their use.

A combination of theory and practice is necessary to grasp the concepts. A recommendation to acquire practical experience without jeopardizing any project, is to implement small pieces of new technology in each project, or by doing small pilots to test the technology.