Using GPS collection to improve GIS data results

Heard Robertson
GIS Analyst
HDR
Atlanta, Georgia

Introduction
In 1996, the City of Atlanta was placed under consent decree by the Environmental Protection Agency, to determine and begin implementing solutions to the environmental problems caused by the city's combined sewer system. It quickly became clear that analyzing the configuration and condition of the existing sewer system would play a key role in determining separation alternatives. It was decided that a GIS-based decision support system would be used to compile and evaluate this information.

In 1999, the city contracted HDR to compile the available sewer and land use information in the 24-square-mile combined sewer area into the decision support system, and to use this information to make recommendations about possible separation or storage solutions to Atlanta's wastewater overflow problems.

The first phase of the project, which began in mid- to late 2000, involved compiling the city's existing sewer information into an ArcView-based GIS System. It was then decided that condition assessments needed to be performed in the field to establish a priority for structure and pipe replacement, repair, and relief. The city used Closed Circuit Television (CCTV) surveys, point inspections, and smoke testing to collect condition assessment information on many of the pipes and structures in the combined sewer area.

During this process it was discovered that the maps the city had on file were, in many cases, extremely inaccurate. The differences between the actual configuration and the GIS configuration made it very difficult to rectify the changes reported by the manhole inspection crews and CCTV crews, thus making it nearly impossible to accurately record the spatial attributes of the system in the GIS. It was decided that collecting positional as well as condition information in the field was necessary to extract an accurate model of the system from the GIS, and that GPS would be the most efficient vehicle for collecting this information.

Tools
Since sewer configurations are contingent on spatial orientation, but not dependent on precise, survey-grade location, it was determined that mapping grade Differential GPS (DGPS) would suffice. The units used were Trimble Pathfinder Pro-XR GPS receivers, which record points with a horizontal accuracy of one meter. These receivers were controlled by Trimble's Aspen software, which runs on a Windows 95+ tablet computer. This software gives GIS technicians the flexibility to project the receiver position on a background map (such as the edge of pavement or building footprints), place structures and lines either with the GPS or manually where GPS signals were not available, and record necessary attribution digitally. Aspen also allows technicians to attach custom dialog boxes to each structure and pipe inspected where attributes could be selected from menus or entered in text boxes. The geometry of the structure as well as the attributes can be exported seamlessly to a number of formats, including ArcView.

Implementation
To establish a clear idea of the condition and configuration of the sewerage system, the city contracted Sanitary Sewer Evaluation Surveys (SSES) in the form of point inspections on the manholes and CCTV surveys on the sewer lines. These crews collected such information as pipe size, shape, and material, as well as any structural, construction, or service defects that were visible, and documented these with digital video and still photographs. This information was then compiled into a database.

HDR's GPS technicians accompanied the City's SSES crews, recording the topology of the system concurrently with the pipe inspection and attribution. The digital inspection data was related to the GIS based on an 11-digit naming convention that was recorded by both the inspection crews and GPS technicians. This number is based on the USGS map quadrant, a random identifier, and a two-digit structural identifier. The data tables were joined together based on this number, and attribution was transferred from the manhole inspection database to the GIS database.

The greatest problem faced by the team was ensuring that the naming convention was observed, and that the name collected by the inspection crews and the GPS technicians correlated on every structure so the information could be electronically linked in the database. The possibility for human error in recording an 11-digit number for each manhole is quite large, so the protocol was modified so that both parties could enter a simplified version consisting of the three-digit random identifier. This number was also painted on the structure lid by the GPS technicians to reduce confusion and to make it easier to relocate the structure in the future.

The sewer system information was supplemented with parcel-based land use information, LIDAR elevation information and aerial photography. It was then analyzed to determine the priority of sewer repair, replacement, and relief, and to produce a conceptual solution to the combined sewer overflow problems that included both partial separation and tunnel storage.

The inventory portion of the project was completed in June 2002. The engineering analysis of the results is anticipated to continue through the end of 2002.

Benefits
A number of benefits were realized in the implementation of the process. The most significant result was that HDR was able to compile more complete, accurate coverage of Atlanta's combined sewer system with less error. The ease of the transfer of the GPS data into the GIS database and the seamless way the attribution information was joined to this database saved our GIS analysts hundreds of hours of data entry.

Since there was an existing GIS database of the city's sewer, it was possible to project this information in the pen computers, and use the GPS to navigate to the location of previously charted structures. This saved inspection crews time in recon and searching for manholes. Another benefit was having employees working on the project who had first-hand experience with the sewer configuration.

The road ahead As GPS and related technologies continue to grow, the possibilities for municipal departments and consultants are also expanding. It is already possible to use Real Time Kinetic (RTK) GPS to get accuracies of less than a centimeter in a matter of seconds. And it is estimated that in three to five years, it will be possible to get this level of accuracy on a device the size of a cellular phone without having to establish any fixed-point correction. GIS technology, GPS equipment, and the software that relates the two are becoming more intuitive, user-friendly and affordable. The result is an increase in the products' ability to span a greater variety of disciplines and budgets.

Advances such as wireless web technology, coupled with internet-based GIS server applications such as ESRI's ArcIMS, are making it possible to access, create, and update enterprise GIS data in real time from the field. This same technology, coupled with a three-dimensional analysis software package, not only makes it possible to update topographic information for watershed analysis and other elevation-dependent applications in real time, but also to use GPS for real-time calculations in construction layout, grading, and quality control.

Other supplemental data collection technologies such as aerial photography, LIDAR, and remote sensing technology allow GIS field data to be compiled from a variety of manned and unmanned sources and increases data analysis capabilities. These various combinations of GIS, GPS, and other available technologies make it possible to create a comprehensive, customized tool for geographical analysis and decision support.

Conclusions
GIS is quickly becoming the standard in data management for utility and public works infrastructure administration, as well as a plethora of other disciplines that utilize data with spatial relationships. In many cases, when data is compiled from a number of hard copy and digital sources that may be in conflict with each other, it is difficult to determine the correct configuration and attach attributes accordingly. On projects where the quality of the deliverable is dependent on exact geographical locations, it may not be possible to resolve these conflicts by conjecture. Simultaneously collecting attribution and location information, whether it be manually, using conventional survey techniques, GPS, or remote-sensing technology is a practical way of ensuring that location and attribution information are correctly correlated. Use of GPS location and tablet computers helped HDR realize an increase in productivity, and produce a product that was superior in quality and functionality.

As GIS use becomes more widespread, the quality and source of data are becoming more of a concern among users. This project is an example of how collecting attribution and location information in the field simultaneously is a practical method of ensuring the integrity of data.

Heard Robertson can be reached at (678) 420-5449 or at throbert@hdrinc.com.