Movable bridges are rare, and they tend to be engineering marvels, but the Cape Fear Memorial Bridge in Wilmington, North Carolina is an unusual case even among those. Rather than swinging or folding, its roughly 400-foot steel center span rises straight up between two towers, high enough to clear the barges and sailboats working their way toward the Port of Wilmington, then settles back onto a four-lane roadway carrying U.S. 17, 76, and 421 across the river.
It has been lifting and lowering since its opening in 1969, and it is now showing its age. More than 65,000 vehicles cross it daily, well past what its designers planned for, and commuters moving between Brunswick and New Hanover counties share the deck with freight trucks running to and from the port. The North Carolina Department of Transportation (NCDOT) calls the bridge safe but becoming functionally obsolete, meaning it can no longer serve the traffic it carries, and it now requires monitoring, inspection, and maintenance on an increasingly frequent schedule. Every lift cycle and every repair adds delay to a crossing with no alternate route. A $1.1 billion replacement is working through federal environmental review, with public hearings scheduled this fall.
Before any of that gets designed, somebody has to measure what is there. That work went to McKim & Creed, a firm founded in downtown Wilmington, just a few blocks from the bridge its survey team would be asked to map. What NCDOT wanted was centimeter accuracy. What it would not allow was a single lane closure.
A bridge you cannot stand on
Bridges are already among the most difficult pieces of infrastructure to access for inspection or scanning. Traditional methods often involve putting workers in harnesses over the side to get close enough to measure. The Cape Fear Memorial Bridge brought more challenges on top of that.
Without a shoulder or bike lane, there was no way to set mobile targets where they would normally go across the span. The heavy vehicle loads the bridge is known for became an obstacle in themselves, introducing vibration that degrades both vertical and horizontal measurement without warning. But the biggest constraint was the most fundamental. NCDOT would prefer not to have lane closures or direct occupation of the deck, and the lift had to stay operational for barge and sailboat traffic passing underneath.
Aerial collection wasn’t the solution either. Drone-grade IMUs could not match what a vehicle-mounted system carries, and from nadir, a lidar sensor may not see a guardrail face at all. The vertical features this project ended up depending on are precisely the ones a flight would have missed. Permitting was the other obstacle, since DOTs generally restrict flight over moving traffic.
None of that changed what was owed. McKim & Creed still had to deliver NCDOT a 3D survey drawing at centimeter accuracy on a structure and a roadway that the crew was not allowed to stand on.
Multimodal creativity
The McKim and Creed Geospatial team, led by Matt LaLuzerne, PSM, national geospatial director, and supported by Trevor Tyson and Martin Taylor, team leads, built the answer around three coordinated data sources rather than a single sensor, each covering the gaps the others left behind.
- Mobile mapping was a key technology for handling the roadway itself. A Riegl VMX-2HA, a dual-scanner system with integrated GNSS and inertial navigation, collected at the posted 45 mph in daytime traffic. Crews ran multiple passes, which served two purposes on a job like this. It built redundancy into the deck surface, and it guaranteed common features appeared in every pass, giving the processing team something to compare one trajectory against another.
- Static scanning was then used to capture the bridge structure and the tender house. Crews set a Riegl VZ-600i at ground positions below the bridge at traditional grade, and again from an elevated position with direct line of sight onto the lift towers, the truss, and the deck edge.
- Control and targets went outside the span entirely, in areas the crew could safely reach and hold. In one case that meant a single parking space adjacent to the bridge tender house, where they could set up and scan while traffic kept flowing overhead.
Keeping control off the span solved the safety problem and created a technical one. Mobile lidar systems hold position by combining GNSS with inertial navigation, and when GNSS degrade (as it can under a steel truss) the inertial solution drifts. Survey targets placed along the route are what normally pull the trajectory back into place. Here there were none, so the correction would have to come from somewhere else.
Building control that was never there
The control network started on the ground. Crews set traditional targets and survey control at grade in the areas where they had clear visibility, then triangulated the static scanner positions from that control. Those positions carried forward to the tender house scans, which gave the team defensible coordinates on the structure without ever setting a point on it.
Holding the mobile data to that framework took a less conventional approach. Riegl’s RiPROCESS defines its own planar features for adjustment, so the team went in and defined features of its own, formatting them into the software alongside the defaults.
The features they chose were deliberate. Guardrail faces became tie-planes, supplemented by light poles and discrete points. Those are not the surfaces a mobile calibration typically relies on.
“Typically you aren’t using that in a mobile calibration,” LaLuzerne said. “You want roadway or flat features. We’re using vertical features to help with along-track adjustment.”
From there the team synchronized the mobile passes, compared the relatively calibrated mobile data against the static-derived planes and discrete points, and applied adjustments to the targeted features. The shifts that came back were minimal.
Checking the work against itself
Validation meant putting the two datasets on top of each other and looking hard at where they disagreed.
The team overlaid the defined plane patches on the static point cloud, then cut cross sections through features they knew well. A guardrail face gave them a clean planar comparison. Light poles gave them something better, because a circular cross section either closes on itself or it does not, which makes it an unusually honest test of horizontal alignment. In the cross-section views, mobile data rendered green against static in purple, and the two run essentially on top of one another.
“Lidar to lidar, multiple passes, different time differences, the whole nine yards,” LaLuzerne said. “We feel very confident in our positioning.”
The final package to NCDOT included the 3D survey drawing, and a calibration report documenting the adjustment process. Accuracy landed at the centimeter level, meeting NCDOT guidelines.
Pre-approval is half the battle
McKim & Creed had to bring the method to NCDOT before fieldwork began and make the case for it, laying out what they proposed to do, why, and how they would prove it worked. The firm had run the approach before and knew it would hold. The pre-planning effort on this project was substantial, closer to how special operations teams rehearse, with plan A, plan B, and plan C worked out before anyone reaches the site. That is the part nobody sees, and it is what let field operations run cleanly inside a schedule built around NCDOT, marine traffic, and bridge tender crews.
The survey crew stayed small, there were two people on static scanning, three on mobile mapping, and five or six in the field overall, with roughly the same number processing the data. The entire process from field to deliverable ran about twelve weeks.
A repeatable answer for hard-to-reach structures
Nothing about the approach is specific to Cape Fear Memorial Bridge, and McKim & Creed hope to apply some of these approaches to other projects in the future. The workflow – especially the repeatable planar adjustment between datasets, anchored by triangulated GNSS control on stable ground – applies anywhere a structure is too dangerous, too busy, or too sensitive to occupy directly. One of the most impressive aspects of the project was not any individual sensor, but the integration of mobile mapping, terrestrial lidar, and survey control into a single defensible dataset.
The outcome NCDOT approved is worth restating plainly. A centimeter-accurate survey of the roadway approaches and portions of the 1969 lift bridge carrying more than 65,000 vehicles a day, delivered without a lane closure, without interrupting marine traffic, and without putting a survey crew into a live travel lane.
“Historically surveyors, survey crews particularly, are put in precarious situations,” LaLuzerne said. “We can avoid this with a little creative technology.”
About McKim & Creed
McKim & Creed is an employee-owned firm with offices throughout the U.S., including North Carolina, South Carolina, Florida, Georgia, Virginia, Texas, Louisiana, Delaware, Ohio and Pennsylvania.
The Raleigh, North Carolina-based company, which was founded in 1978, specializes in civil, environmental, mechanical, electrical, plumbing, and structural engineering; industrial design-build services; airborne and mobile Lidar/scanning; unmanned aerial systems; subsurface utility engineering (SUE); and hydrographic and conventional surveying services for the energy, transportation, federal, land development, water and building markets.
