How one software linked design and construction operations on the world's longest national road.

The Trans Canada Highway is the world’s longest national road, extending across the entire breadth of Canada for 4,860 miles (7,821 km). From its western terminus in Victoria to the Alberta border in the Yoho National Park, this scenic highway is maintained by the British Columbia Ministry of Transportation and Highways (BCMoTH). In June of 2000, the Ministry completed construction of a new interchange at the intersection of Highway 97 near the city of Kamloops. And one software program eased the process.

With the project at hand, and possessing an aging software system, the Ministry decided to implement a new system for its surveying and highway design operations, beginning with the Trans Canada Highway project. It also sought a single software solution to integrate its survey and design processes with its construction and geotechnical engineering operations. After some evaluation, the Ministry selected the software system of CAiCE Software Corporation of Tampa, Fla., for a pilot program to automate its construction operations. The pilot program required the building of 4 km of four-lane divided highway, 3 km of two-lane highway and 6 km of frontage roads and ramps.

Prior to the implementation of CAiCE software, the data processing methods the construction group used required a great deal of hands-on data entry and station-by-station, surface-by-surface and point-by-point editing. There was no real software integration, and information sharing between software systems was accomplished via in-house file conversion utilities. The process was very rigid and unforgiving, making error correction unpleasant and laborious. By implementing a single software package for survey, design and construction, BCMoTH hoped to streamline the entire flow of information from conception to finished construction. They also hoped to create an environment in which construction information would flow back smoothly to the design group in the event of project revisions. CAICE software helped with these goals.

The Pilot Project Begins

To meet the needs of the Ministry’s construction group, CAiCE had to address several areas. These included estimating quantities of topsoil stripping and removal of unsuitable materials, calculating periodic excavation volumes for progress payments to the contractor, providing estimates of rock removal quantities and comparing as-builts to the original design models. Selecting the same software for both the design and construction groups was a key step in meeting these goals.

Once the design was completed and a contractor was selected, the construction group received the CAiCE project database. The collection contained the original survey data, existing ground digital terrain models (DTMs), complete construction plans and roadway design cross sections taken at even 20-meter stations and other special stations of interest. By this time, however, the original topographic survey was several years old. So, the construction group re-surveyed the existing terrain, replacing the original survey data and ground models, and ensuring that the models were completely updated. Special attention was paid to surveying the locations of unsuitable materials and rock outcroppings.

Many of the software operations and capabilities needed for the construction operations were similar to those the survey group used. Both groups must import field survey information from data collectors, and edit the points and chains to build accurate digital terrain models for different types of surfaces. Construction survey crews were already proficient in string and cross section survey techniques and only needed to make minor changes to their data collection methods for the new software.

One major difference between survey and construction applications is that while the survey group develops a static model of the existing terrain conditions, the construction engineer must model surfaces that are constantly changing as the construction progresses. Indeed, one of the principle tasks of the construction group is to create models of the same site at different points in time that can then be compared to calculate progress payments for the construction contractor. This creates some data management issues that can be overwhelming if not organized properly. Early in the project it became apparent that a procedure had to be developed to deal with the large number of survey files that were being created, and to differentiate between the datasets collected by separate re-surveys. Good communication between field and office staff was essential.

As an example, one group of survey files may measure the topsoil stripping removed within a defined chainage range along a specific alignment during one month. Within the overall project database, that group of data must be easily extractable to build a DTM representing the volume of stripping for that alignment for that month. Other groups of survey files within the same project may represent further stripping at different chainages along the same alignment for different months. Once all stripping is complete, it may be desirable to build a single DTM surface to determine the overall stripping volume for the entire alignment. Meanwhile, other surveys are being done to measure normal excavation and rock excavation on the various alignments.

To keep all of the data properly organized, strict file naming and data coding conventions were developed and rigidly followed by the surveyors and computer operators. Fortunately, CAiCE offers a number of user-definable attributes for data objects that separate different types of materials and groups of survey data. As it turned out, these standards were the key to ensuring a smooth flow of usable data into the automation process.

Construction automation also involves many other operations unique to the construction group. To a large extent, it relies on the ability to interactively edit points and surfaces in cross section form and to apply editing operations globally across multiple stations. Extending surfaces to close gaps, trimming overlapping surfaces, adding new surfaces during project construction and combining surfaces to form closed areas are daily occurrences. These cross section utilities differ greatly from those used by the design group, which are based on modeling typical design sections using templates. Also, there are a number of report utilities specific to construction operations, such as volume tabulations and layout reports for slope staking and gravel grades. CAiCE worked closely with the Ministry to develop the needed tools.

Figure 1. Perspective of rendered normal terrain and rock outcroppings.

Steps Made Easy

CAiCE has the ability to visualize survey data and surface models using colors, texture bitmaps and aerial photographs to render the surfaces. CAiCE also makes it easy to step around and drive through a project, viewing the data from different perspectives. The visualization capabilities of CAiCE proved very useful in field-office dialogues, and the ability to see the surveys in rendered three-dimensional images has helped significantly in reducing errors. Figure 1 shows a three-dimensional perspective of the terrain using texture bitmaps to render normal terrain and rock surfaces.

The next step in the process was to use the re-surveyed existing terrain to rebuild the roadway design models. Cross sections for the original ground (OG) and rock surfaces were taken from the DTMs and merged with the design grade surfaces. Due to the rebuilding of the OG surface, there were many instances when the original design catch points no longer closed exactly on the existing terrain. CAiCE helped to streamline the recalculation of slope stake points by providing utilities to globally trim the overlaps and extend the catch slopes where necessary. This eliminated the need to edit the catch points on a station-by-station basis.

Figure 2. Uniform topsoil stripping.
During the early phases of construction, the contractor began stripping topsoil and removing unsuitable materials (see Figure 2). In most cases, each stripped area was resurveyed, enabling the Ministry staff to build a DTM representing the stripped surface. Cross sections of the stripping areas were then added to the design cross sections. In some cases, a stripping survey was not obtained, and the contractor and Ministry representative agreed upon the boundary by stations and offsets, using a uniform stripping depth. This required the computer operators to build surface models by extracting the original ground surface within the agreed upon boundary, then dropping it vertically by the stripping depth.

As is often the case on construction projects, there was considerable deadline pressure for compiling end-of-month progress payment reports. The monthly deadline crunch was primarily due to the number of survey files coming in from the field and the time it took to completely process them. One shortcut that proved useful was the ability to calculate volumes between two surface models. This tool was particularly helpful for estimating stripping quantities, but was also applied to other excavation volumes. It produced quick results that were sufficiently accurate for estimating purposes, and adjustments were made after the more accurate final payment quantities were developed.

A considerable amount of editing was required to calculate the final payment volumes for stripping. At many stations vertical links had to be inserted to close between the stripped surface and the original ground. Since the Ministry only pays for soil removal within the design construction limits, stripped areas beyond the slope stake points had to be clipped out. There were also areas where the stripped surface was actually above the originally surveyed terrain. These also had to be removed from the payment quantities. CAiCE addressed these issues by providing visual, interactive, easy-to-use tools for editing the surfaces.

Figure 3. CAiCE cross sections defining material end area envelopes.

Testing the Work

Once the actual grading of the roadway commenced, the Ministry’s construction group monitored the progress by taking monthly re-measure surveys of the site. Each survey was imported into the CAiCE project, edited and used to construct a new surface model depicting the current state of the job.

“CAiCE’s three-dimensional visualization, navigation and drive-through tools proved invaluable by enabling us to see what the site actually looked like to the computer as the construction developed,” stated one Ministry engineering assistant.

Each re-measurement was used to create new cross section surfaces that were merged with the design, OG, rock and stripping surfaces. At this point, the editing of the cross section surfaces started to become somewhat complicated. The re-measure surveys did not always yield complete cross sections, and the surfaces were often left hanging. Good field-to-office communication and graphical visualization were vital so the computer operators knew how the surfaces should close to form end area envelopes.

For this project the Ministry had to account for two different types of materials. Type-D material represented normal earth excavation or fill, and Type-A material represented the rock excavation, which is far more expensive. The re-measurement surveys were carefully performed so these materials were accurately tracked. Special data attributes such as feature codes and zones were used to differentiate between materials, and considerable cross section editing was needed to combine and extrapolate the surfaces correctly.

CAiCE calculated earthwork and material volumes from cross sections by analyzing the end area envelopes formed by the different surfaces. An earthwork classification table lets users define the names and types of materials, based on which surfaces are above and below each area. The end result of the re-measure surveys and the editing was a set of cross section surfaces that defined end area envelopes for topsoil stripping, incremental excavation and embankment quantities from month to month, rock excavation, and five different asphalt, base and gravel paving materials. An example of this is shown in Figure 3. The earthwork tables were flexible enough to handle everything from single isolated material quantities to very complex, multi-material end areas.

CAiCE also provided tools for refining volume quantities for a number of special situations. For example, areas of rock excavation and topsoil stripping began and ended at odd stations that fell between the stations in the cross section file. It was necessary to extrapolate these materials to their actual boundaries to obtain accurate quantities. There were also bridge openings where the volumes had to be zeroed out, and changes in shrink and swell factors.

By the time the pilot project was completed in June 2000, the Ministry had achieved the dual purpose of proving the viability of the CAiCE software for their construction operations and refining the methods, procedures and workflow for making the most efficient use of the software.

“I can see the ground lines as if I were standing in the field. Once the field information is received, we can instantly see a three-dimensional image of the site, with the ability to toggle surfaces on and off to visually show a before and after analysis,” said one computer operator.