Surveyors in North Carolina needed to collect data on rail centerlines and corridor boundaries on 317 miles of track that’s 150 years old. Mobile scanning met their needs safely and efficiently, and its resulting data will help support future monumentation.
Locating rail centerlines and corridor boundaries more than 150 years after they were built isn’t easy, but that’s exactly what the North Carolina Railroad Company (NCRR) needed to do. NCRR built its railroad in the 1850s, with a 200-foot boundary corridor extending 100 feet on either side of the rail centerline. Today the track spans 317 miles, and each day it carries 50 to 60 Norfolk Southern freight trains and 10 Amtrak passenger trains.
In the century and a half of the railroad’s existence, development has significantly encroached upon NCRR’s boundary corridor. “We needed a much better idea of where our original centerline and corridor boundary lie,” said Kristian Forslin, GISP, GIS coordinator with NCRR. “Having this information allows NCRR to better manage our corridor as well as provide a firm basis for future rail engineering projects.” Coordinating available track time with optimum GPS sky view while obtaining survey-grade accurate geometry of the existing tracks for 317 miles proved to be a very challenging task.
NCRR first tried a GPS mobile solution to collect the data. “Our initial test with the GPS approach presented a level for potential failure in a number of areas that NCRR was not willing to accept,” said Forslin. NCRR then decided to pursue mobile scanning. “Mobile scanning technology promised to be far superior based on speed, reliability, and its thoroughness in one pass,” he said.
NCRR commissioned the Raleigh, North Carolina-based firm of McKim & Creed to collect accurate existing rail centerline locations and produce coordinate geometry of all tangents, curves, spirals, etc., of those centerlines. Because the track is so heavily used every day, surveyors had just one opportunity to collect the centerline data. There could be no equipment failures, missed opportunities, or do-overs. The data will be used to support eventual monumentation of the NCRR corridor boundary.
Mounting and Calibrating the System
McKim & Creed’s mobile data collection (MoDaC) system uses two GPS units, an inertial measurement unit (IMU), and a distance measurement instrument (DMI) that work in concert with two lidar sensors and three digital cameras to collect survey-grade accurate positional data. We mounted the system on NCRR’s hi-rail vehicle—a Chevrolet Suburban retrofitted to travel on tracks—using a welded aluminum frame. Once the system was mounted onto the hi-rail vehicle, we rechecked the calibration by scanning a nearby building and made the necessary dimension adjustments between the DMI and IMU units.
Mission Planning and Coordination
With its abundance of conifers, North Carolina lives up to its nickname, “Land of the Longleaf Pine.” In many areas along the NCRR track, the canopy from tall pine trees limited the GPS sky view to satellites. The team had to schedule data collection during optimum GPS conditions when the most satellites and the best geometric positioning were available.
The IMU and DMI units on the MoDaC system helped tremendously when GPS signals were temporarily lost or blocked. We coordinated the optimum GPS conditions with available track time so the team could minimize the amount of time we were on the rails and avoid interfering with the regular train schedules.
Safety
“Safety along the NCRR line is a primary objective for our company. Rail corridors are for trains, not for pedestrians,” John Atkins, chairman of the NCRR board of directors, was quoted as saying in NCRR’s 2010 annual report. The same is true for McKim & Creed. Safety is a primary objective for our company, and we understand that rail corridors are for trains, not for surveyors.
All personnel who accessed the property were certified in both e-Rail Safe and Roadway Worker Protection. Surveyors collected data primarily from inside NCRR’s hi-rail vehicle, accompanied at all times by representatives from both NCRR and Norfolk Southern.
Ground Truthing
NCRR needed accuracies within a 0.10-foot tolerance. To ensure this accuracy the surveying team conducted ground truthing conventional surveys at approximately half-mile increments along the corridor. We located existing pavement markings near at-grade crossings such as stop bars and edge of travel-way lines. Temporary v-shaped targets were set and controlled in areas where pavement markings were not present.
All ground truthing points were obtained using RTK-VRS collection from the North Carolina Geodetic Surveys network while avoiding “fouling” the tracks. These points were used to adjust our lidar data to achieve the required accuracies, especially in the poor GPS areas that were impaired by blockages.
Processing the Data
The collected data was initially processed to a Universal Transverse Mercator (UTM) coordinate base, which ensured that all scan data, GPS data, and imagery were collected accurately and thoroughly. Once the raw data was processed, it was computed horizontally to the North Carolina State Plane Coordinate System, NAD 83, and vertically to the North American Vertical Datum, NAVD88.
Various best-fit computation routines were incorporated to provide true geometry of the alignments. McKim & Creed delivered our best-fit compilation of the existing track centerline geometry for the entire 317 miles in MicroStation, ArcGIS, and GEOPAK formats to be used by the surveying community for future right-of-way monumentation needs.
Images and Videos
Additional deliverables included dual imagery as the project was scanned and a forward-facing completed video (See PSM’s website for a link to the video). The video was created with two 5M-pixel cameras—one facing forward and one facing toward the rear—that are part of the company’s mobile scanning system. The camera placement allowed the surveyors to photograph both approaching and departing views of items of interest along the track.
Also contributing to this video production was Point Grey Research’s Ladybug3 camera system, which runs independently of the MoDaC system and uses six 2M-pixel cameras that fire simultaneously to produce one 12M-pixel image. Creation of this video involved taking the spherical images, flattening them to a screen view, and creating a movie from the images. Rail personnel can use the video to view signs, switches, derails, bridges, road crossings, etc., along the drive path.
In addition to the dual imagery and video, McKim & Creed created a KMZ file that opens Google Earth, zooms in to the geographic location of the route, and displays our photos of that particular location.
What’s Happening with the Data
NCRR is using the data to support monumentation of the corridor boundary. When that is complete, NCRR will fine-tune its GIS with the newly acquired centerline data and monument locations.
“Although a full-blown lidar data-acquisition deployment may be seen as overkill for our project, it was the best way to quickly get the accurate centerline data NCRR needed,” Forslin said. “This information will be a great asset to surveyors who will have a significant amount of new control to choose from as well as a reduced need to obtain right-of-entry permits for work adjacent to the corridor.”