No Roads Untraveled
Wisconsin’s Department of Transportation (WisDOT) oversees 56,153 square miles of land including 982,400 lake area acres (1,535 square miles). With eight districts performing more than 100 projects per year, our survey needs are high. And with a geography that combines woods and water, hills and valleys, WisDOT’s staff has sought new and efficient ways to survey the diverse state.
At WisDOT since 1982 and head of the Geodetic Unit since 1989, I’ve been a strong proponent of GPS use for WisDOT’s surveying operations. As a supporter of GPS since its early days, I realized it would be a powerful tool for accelerating project delivery, reducing costs and maintaining survey accuracy of transportation projects. Today, GPS is key to the new direction our department is taking.
Early GPS Use at WisDOTAs early as 1985, WisDOT used GPS to establish project control and to fix photogrammetric targets for design survey work. We initially hired contractors for GPS work; in 1989, WisDOT purchased its own equipment. GPS was first used in static baseline observations to bring in horizontal control from National Geodetic Survey triangulation stations far cheaper than conventional methods, even over greater distances.
The biggest challenge at that time was that it took a dedicated effort to use GPS. Because the satellite constellation was limited, we had to go out at times between midnight and 5:30 a.m. so we really had to want to use it. It was also difficult to get adequate training. In some cases, we knew more than the people from whom we rented. After a while, I started saying, “Give me the equipment and I’ll figure it out.” Those who went through the early years of GPS implementation know what a cakewalk it is to use GPS today!
GPS quickly won the support of the district survey staff. The initial GPS “proof of concept project” was completed in three to four days using GPS, instead of the usual four weeks, and accuracy was comparable to conventional tools. Estimates determined cost savings over conventional surveying at 65 percent of cost, or $100,000 for this initial project.
RTK Evaluation using UDNIn 1992, WisDOT used GPS differential positioning techniques to establish the state’s High Accuracy Reference Network (HARN). As a result, WisDOT developed a densification plan for establishing geodetic networks (User Densification Networks (UDN)) at various accuracy levels using Federal Geodetic Control Subcommittee (FGCS) GPS Survey Standards that we modified. By changing the FGCS standards and specifications, we attempted to develop realistic criteria for data acquisition, data analysis and data submittal of GPS surveying for WisDOT. Wisconsin counties interested in densifying their areas were given assistance by WisDOT, provided they followed WisDOT standards and specifications. Today, these UDNs are being incorporated into another new but similar initiative called the Wisconsin Height Modernization Program (WI-HMP). WI-HMP utilizes similar station spacing as the UDNs for horizontal work; however, extensive leveling is performed to establish orthometric heights.
After using GPS for some time, WisDOT wanted to access the capabilities and advantages of Real-Time Kinematic (RTK) GPS to position points with 1-2 cm accuracy in the field.
To determine if statewide RTK use was feasible, we needed to find out whether we could obtain RTK solutions from Wisconsin County UDN (or WI-HMP) at the 2 cm accuracy (3D) 95 percent confidence level. If so, we could then use RTK with assurance for most of our applications.
We performed UDN/WI-HMP evaluations at four locations throughout the state with Trimble GPS Total Station 4800 units (Trimble, Sunnyvale, Calif.). Our results showed that RTK use was possible when using UDN/WI-HMP reference stations located approximately 6 km apart. We were confident that required accuracies could be achieved using RTK for project control, photogrammetric targets, section corners, right of way monuments and other applications. Results showed little difference between conventional GPS procedures and simple RTK procedures using averages from two reference stations.
Since then, WisDOT has used both RTK and static GPS to provide nearly all our survey needs. Today, WisDOT has 16 Trimble GPS RTK systems, at least 72 total stations and one aerial exposure/in-flight navigation GPS. We use the Trimble TSC1 data collector with Trimble Survey Controller software and Trimble Geomatics Office software, which allows complete interoperability with all total stations and GPS equipment.
Project ControlWhen we started using GPS, project control had been established by a central office geodetic survey staff utilizing Trimble 4000 SSi GPS receivers. We used budget initiatives as an opportunity to simultaneously achieve cost reductions and train district staff surveyors. With implementation of RTK by district staff and a reallocation of central office staff, project control is now primarily established through consultants.
One reason WisDOT chose RTK was to increase the district’s ability to take a project from start to finish without relying on the central office for project control. An important aspect of project control is to first determine how good existing control is in order to tie into existing projects.
Wisconsin counties with UDN/WI-HMPs at 6–8 km control spacing typically do not require additional project control because the RTK system can be used routinely to adequately perform other survey tasks in the project area. However, counties that are not densified will need additional project control established to facilitate RTK use. Under these circumstances, project control is established at 6–8 km surrounding the project.
Photogrammetric TargetsA strong point of RTK is the ability to rapidly and accurately determine the position of photogrammetric targets. Previously, positions on photogrammetric targets were established using conventional total stations. We then progressed to positioning via rapid static procedures using GPS. Initially, the central office used a central survey crew operating GPS receivers to establish control, but now all districts have purchased and use the RTK systems.
With RTK, the process uses county UDN/WI-HMP stations surrounding the project area for control. These stations are established at 6–8 km intervals with an accuracy of 1 part in 250,000. For the procedure we set up the base receiver over one known UDN/WI-HMP station and checked another UDN/WI-HMP station with the rover. If the misclosure is within 1 part in 250,000, the rover then occupies each unknown photogrammetric target for 3–5 seconds. After each unknown photogrammetric target is occupied, the rover then occupies another UDN/WI-HMP station as a check. We then move the base receiver to another UDN/WI-HMP station and repeat the procedure through other UDN/WI-HMP stations and the unknown photogrammetric targets. Subsequently, we compute and verify the results in the field with Geomatics Office software.
If UDN/WI-HMP stations are not present in the project area, project control is established surrounding the project area using static, fast static, or both procedures. Personnel in the central office geodetic surveys unit or consultants often perform this process. Following the establishment of project control, district survey staffs will determine horizontal positions of the photogrammetric targets using RTK as described above.
If a photogrammetric target cannot be occupied by GPS then a pair of stations are established nearby and the target is surveyed conventionally with a total station.
Section CornersRTK GPS enables WisDOT districts to do section corner surveys independent of the central office. In most cases, project control serves as the basis for establishing control on section corners using RTK. RTK GPS enables WisDOT surveyors to occupy section corners more quickly and with the same accuracy as a total station. The exception to this approach is when section corners cannot be observed due to natural or artificial blockage of the satellite signals to the receiver antenna.
In these instances, we set two to three control points nearby and determine the section corner coordinates with a total station. Depending on tree growth density, 10–50 percent of section corners on a project would have to be surveyed conventionally. In many, if not all aspects, section corners are surveyed in a manner similar to photogrammetric targets.
Another benefit of GPS is when section corners are not located along fence lines or other obvious indicators of possession. In these cases, we use GPS as a search tool by taking the best estimate of the corner position as a waypoint and using GPS to navigate to it, speeding up the process of finding it. Once found, its accurate coordinates can be determined using RTK GPS.
Topographic Mapping/Digital Terrain Modeling (DTM)Normally, when WisDOT wanted to replace or build a structure, we did it photogrammetrically, using photos to construct the base map. For example, on small structures such as bridges we map an area approximately 500 ft from each end of the bridge along the road by 100 ft each side of the centerline and an area approximately 500 ft up and down stream by a dimension determined by terrain on each side of the stream.
Mapping accuracy requirements vary depending on the feature being surveyed and whether it is in or close to the right of way. DOT requirements for line and natural features are typically horizontal accuracies of 0.10 ft and 1.00 ft, respectively. Elevation requirements on pavement and ground typically are vertical accuracies of 0.04 ft and 0.10 ft, respectively.
Now with GPS, the surveyor produces not just the control, but the data for compiling a map. Since the surveyor is actually in the field, he or she can pick up additional subsurface features as well as the crucial area under the structure that cannot be easily picked up by aerial methods. This is particularly true for locating utilities in and near the project area with a horizontal accuracy of 0.2 ft.
The normal procedure for collecting field data is to set up the RTK base receiver over a known station with horizontal and vertical positions. The rover checks at least one other known station and then collects topographic information. At the survey’s end at least one known station is occupied to verify accuracy. Sometimes two rovers are used to expedite the survey. In obstructed areas, a pair of stations is established using RTK and the feature is surveyed conventionally with a total station. Data collected from RTK GPS is processed through Geomatics Office. Data files are converted to Survey Data Management System (SDMS) format and then imported into Computer Aided Civil Engineering and Surveying Visual 2000 SE software for review (CAiCE Software Corporation, Tampa, Fla.).
Similarly, before we had RTK GPS, when rehabilitating or reconstructing road intersections, stereo imagery would be acquired for up to five miles of the road on each side of the intersection. The maps would then be compiled from the aerial photos.
Now with RTK, surveyors can collect data to produce a topographic map and DTM of the intersection with a shorter elapsed time than with aerial photogrammetry. An added benefit of RTK is that designers don’t have to contend with less-than-desired accuracy from aerial mapping in areas that are hard to compile with photogrammetric methods. Additionally, with RTK you can visually see and pick up additional information you can’t get from photos.
We found again that districts with either RTK or total stations can do this more efficiently and effectively—and on their schedule—than the central office’s personnel and photogrammetric equipment. With 80–100 photogrammetric projects a year in the state, many smaller projects weren’t getting done for three to four months. Now the districts can do these jobs themselves.
Even where photogrammetry is used for topographic mapping and DTM development, WisDOT has found that conventional total stations are required 50 to 70 percent of the time to complete the work in areas with obstructions such as tree coverage, towers or tall buildings.
Staking Right of WayIn addition to the obvious benefits of staking highway alignments, RTK also makes staking right of way more efficient. Staking right of way doesn’t just include setting points to demarcate the property to be used by the state, though that is the largest part of the process. Other tasks include staking out tentative alignments and corridors to enable appraisers and right of way agents to negotiate with landowners and staking locations of temporary easements during construction for areas of material storage and the like.
Staking right of way may be needed two to three times before final staking depending on what occurs during the acquisition through the final construction project phases. A typical field procedure using RTK is to occupy at least four section corners surrounding the area to be surveyed and then navigate using station and offset values in Geomatics Office’s Roadlink processing option. In at least one district, station and offset is primarily used since they can set intermediate stations every 500 ft along the tangent if there are obstructions. Again, if obstructions are problematic, at least two stations are set with RTK and then the right of way point can be surveyed conventionally with a total station.
Since introducing RTK, WisDOT has reduced personnel hours up to 70 percent compared to what it normally took to perform these same tasks using conventional equipment and procedures. Prior to RTK, a three-person survey crew performed right of way stakeout. Today, a two-person survey crew accomplishes the same tasks with RTK. On rather small projects a one-person survey crew could perform the duties; however, typically a two-person crew is needed to maintain safety and efficiency of other activities.