Tracking the Railway with GPS
South Florida has a population base that has exploded in recent years. As with most regions experiencing the challenge of rapid growth, transportation is a major issue for the local and state governments. In South Florida, a fundamental component of public transportation—the Tri-County Commuter Rail Authority (Tri-Rail)—helps to alleviate the burden on the local roadways. Using existing railways, Tri-Rail has been providing commuter transportation in Miami-Dade, Broward and Palm Beach counties for the last 13 years. The rail is shared by Tri-Rail, Amtrak and CSX Transportation (CSXT). To provide better service to its existing and new customers, the Florida Department of Transportation (FDOT) initiated a double tracking improvement along the South Florida Rail Corridor.
Tri-Rail LogisticsThe project, dubbed the Tri-Rail Double Track Corridor Improvement Program Segment Five Project, is a massive undertaking to add 44 miles of additional track to the 72 mile rail corridor. As with any major construction project, much of the initial effort belongs to the surveying profession. After the preliminary surveying and engineering was completed, the design/build contractor (who constructed the actual improvements) led by his own surveyors performed a more detailed survey of the existing corridor. One of the requirements of the design/build contractor was to provide data to Tri-Rail and the Florida Department of Transportation (FDOT) in State Plane Coordinates (SPC) using the North American Datum of 1983 (NAD 83) with the 1990 adjustment (NAD 83-90).
Inside the rail corridor are a number of concrete monuments placed approximately every 1,500 feet. The location of these monuments was recorded on FDOT right of way maps provided to the design/build contractor. Unfortunately, these maps were created in the late 1980s and the coordinate system used did not have the 1990 adjustment.
In order to assist the future survey teams of the design/build contractor, the monuments needed to be recovered and NAD 83-90 coordinates provided for them. The goal was to recover or replace the existing FDOT right of way control monuments within the rail corridor and place them in the current coordinate system. Reference points for the monuments in the current and future work areas needed to be verified in the field. We at Tri-Rail worked in three distinct sections in three different counties with one section falling in two counties. Our final product would have the monuments defined with new coordinates and related to the calculated right of way as determined from the FDOT maps. The information was provided in a CADD format using Autodesk Land Desktop (Autodesk, San Rafael, Calif.) and MicroStation J (Bentley Systems Inc., Exton, Pa.). We created right of way linework in conjunction with surveyors at DMJM+Harris, an engineering firm in New York, N.Y., which also provided another independent check of the calculated right of way.
Equipment and PlanningIn planning the project, the Project Management Consultant’s Project Surveyor, Douglas Davie, PSM, of Craven Thompson and Associates (CTA, Fort Lauderdale, Fla.), took several factors into account. The first issue was the small window of time available to accomplish this task. In fact, issues involving the use of CSXT Railroad flagmen needed to safely access the corridor made limited time an even larger factor further into the project. CTA required these flagmen for the other major factor Davie had to consider—safety. With hundreds of tons of train barreling toward them at speeds of up to 70 mph, survey personnel needed to be ready to move away from the tracks at a moment’s notice. Constant communication with the CSXT flagmen had to be maintained while working in the South Florida Rail Corridor.
After weighing the factors involved, it was decided that the most practical way to accomplish the job would be to use Global Positioning System (GPS) equipment. CTA has Trimble GPS equipment (Trimble, Sunnyvale, Calif.) in its inventory and decided it would be put to use on the Tri-Rail job. The time it would take to traverse the length of the project areas (one of which was over 30 miles long) proved unacceptable for the time frame of the project. GPS equipment was also considered safer for the field personnel. There would be no tripods to carry (tripods are a major factor in train accidents as field personnel tend to carry them on their shoulders, blocking their peripheral view of oncoming trains). Time near the tracks would be reduced due to increased production, and the better mobility of GPS equipment would make the user more likely to move out of the danger area, as the time to stop and restart a session is negligible as opposed to trying to store that last shot with a total station and data collector.
The Network PhaseThe control survey was divided into two phases. One phase involved building the external network, what we referred to as the Network Phase. This phase involved the research and field recovery of existing GPS control points. After the points were identified in the office, field crews took the existing location sketches or descriptions and attempted to find those points in the field. At times a GPS receiver was needed to find an actual point. Even in an autonomous mode a GPS receiver should lead a person to within 10 meters of a point. On occasions where the “To Reach” descriptions were too vague or wholly incorrect, the capability of GPS precision proved invaluable to field personnel. If a point was recovered, we evaluated its condition and the surrounding terrain to determine if it was truly usable. It was vital that the person recovering these points understood how the GPS system works and what parameters to look for—and to make the same notes in their field books or recovery sheets. This way those in the office understood whether a point should be included in the network. When dealing with so many points over such large areas you have to rely on the field evaluation from the survey crew actually performing the site visit. At least twice the planned network needed to be readjusted when field sessions were underway because of miscommunications on the viability of points and vague obstruction sketches from recovery personnel. Luckily, these were in areas where other recovered points existed and could be shifted with minimal delay.
Once enough points were recovered, the preliminary network was planned and executed. The larger the “planned” network, the more “preliminary” it becomes. GPS control points can be lost due to field conditions (road or site construction and/or monument disturbance). Also, problems in the session data collected (high interference, procedural error) will cause a readjustment of the final network you will end up with. Because of the size of this particular project and the timeline involved, there was a level of flexibility built into the project from the beginning by identifying a denser distribution of control than would be necessary for the eventual project network.
The Corridor PhaseThe other phase of our control project involved occupying monuments within the actual rail corridor, what we referred to as the Corridor Phase. This phase involved working within the railway and involved close interaction among survey crews and the CSXT flagmen. The flagmen served as the eyes and ears of the survey crew and provided early warning of all approaching rail traffic. Primary means of communication were through cell phones with a radio as backup. While the network phase would normally utilize multiple crews and vehicles, the corridor phase typically involved a single two-person crew. This allowed them to build a good working relationship and a familiarity of all the terminology and procedures used when working in the corridor. (Special classes and certification are necessary before personnel are allowed to work in the corridor.)
The areas where we provided our field control involved three distinct regions in three separate counties. The largest region was in Palm Beach County and involved seven separate construction areas. In each county these work areas were adjacent, but our scheduling was based on the timeline for the overall Tri-Rail Double Tracking Project rather than location of the area. As a result of this, our fieldwork did not involve a smooth progression from north to south, south to north or even one phase to the other. The reason we referred to our phases as Network and Corridor rather than 1 and 2 or A and B was simple; either phase could be done first and this sequence varied in the different counties. In fact, during the course of our work in Palm Beach County, we ended up switching from Network phase to Corridor phase and back again. The main constraint involved the CSXT flagmen who needed to be scheduled far in advance and issues of flagman availability altered our field schedule further.
Our initial fieldwork was delayed due to contractual negotiations. This meant that once we resumed work, the timetable was even tighter than before. One benefit from the schedule change was that the additional time was used for the field personnel to become familiar with the equipment, and it allowed us to adjust any network changes with additional sessions. We were also able to refine our field procedures and identify any kinks in the process both in the field and the office. The lessons learned in these initial GPS field sessions greatly enhanced our productivity when the project commenced.
Our strategy with this control survey was to establish various primary control points along the corridor at intervals of approximately 2.5 to 3 miles. These points were set in areas identified in the office as viable locations for GPS control and confirmed in the field by the survey crew. These “CTA” points needed to be open, easily accessible and reasonably adjacent to, but not within, the South Florida Rail Corridor. Using this control we could tie in the monuments inside the corridor without spending valuable time traveling various distances to control points of varying degrees of safety. This also minimized or eliminated interference with ongoing civilian activity (setting up our equipment near roads or in front of a business or residence).
County Challenges, Each county involved in the project presented different challenges. In Palm Beach County we found numerous road construction projects underway that not only impacted travel time but also denied us the use of several control points. Because of the timetable of the work in Palm Beach County, adjacent areas of work could easily be merged into one network. The time lapse between field sessions meant that several GPS control points could not be common between the adjacent regions because they had either been destroyed by construction or were no longer usable. This forced us to use longer baselines and additional field sessions to merge the network for these areas. In Broward County we found numerous GPS control points but experienced one or two locations along the corridor that precluded the use of GPS due to man-made obstructions. When GPS could not be used, we used traditional surveying methods, in some cases, Leica Geosystems (Atlanta, Ga.) TCR 703 and TC 800 total stations. If the railway control resulted from total station measurements, scale factors were utilized to produce the needed State Plane Coordinates. Miami-Dade County proved the most difficult to find GPS control within. The smallest region of work fell within Dade County. A disproportional amount of time was spent searching for usable GPS control points compared to fieldwork of both network sessions and corridor locations.
Once the GPS control work was completed, CTA also had the job of recovering reference points for the control and, if such points could not be recovered, referencing the monument. For this work we returned to conventional methods of field procedures. As we expected, most of the reference points were still intact. This is fortunate because the time involved for referencing these monuments would have greatly exceeded the time to locate the monuments themselves. The monuments were FDOT concrete monuments with brass caps. Their condition varied along the corridor but for the most part they seemed to hold up very well. We used GPS to locate the initial monuments. Although the coordinates we had were from a pre-GPS era and the network was not completely tied down yet, we presumed the equipment would be able to fix the monument location within 2 to 3 feet. In the field, the crews reported a shift of less than a foot (which was consistent with the expected shift from NAD83 to NAD83 with the 90 Adjustment). The fact that our equipment was able to lead us to the points before we tied them into our control network is perhaps an overlooked benefit of using the GPS. The lines of sight from one monument to the next could easily be obscured by vegetation growth that did not exist in the original survey.
In reflecting upon the work done and the lessons we learned, the project was a success in many ways. The productivity of the field crews grew tremendously as the survey advanced. Much of this was due to the adaptability, hard work and professionalism of our field personnel who adjusted from their normal tools (total station, steel tape, etc.) to GPS. Once they saw the time and effort savings, they openly embraced not having to endure the long setups involved with traditional surveying methods. It was not long before the office was struggling to keep up with field production.
There are many jobs where GPS is found to be unusable or impractical. However in this case, due to the size, nature and constraints of this particular survey, it proved invaluable for CTA and all the parties involved. The key for GPS work is simple: planning. An extra hour of planning in the office can easily save four hours in the field and cannot be neglected. GPS is just a tool and can be managed or mismanaged accordingly. Thorough planning, coordination and, above all, flexibility demonstrated themselves as vital components for success in this surveying effort for the Segment Five Project.