The technology may still be young, but it is teeming with potential. While typical uses include aerial mapping, emergency management and disaster assessment, inspection applications are of great interest to the maintenance and asset management industry.

As yet, unmanned aerial systems are not widely used in bridge inspections, but given the number of bridges across the USA and the requirement in many cases that they be inspected as often as once every two years, a broader use of UAS for bridge inspections could result in greater efficiency and worker safety as well as cost savings for bridge owners. 

UAS technology has received considerable traction as an inspection tool in a variety of sectors and its capabilities are now being more thoroughly unlocked for bridge inspections. Through past experience on projects using UAS, it is believed this technology can play a more extensive role in bridge inspections and provide a number of key benefits. UAS was deployed to assess fire damage on the Liberty Bridge in Pittsburgh, Pennsylvania — one of the region’s most important spans — and a more extensive bridge inspection pilot project was completed for the Wisconsin Department of Transportation in 2016.

Because of their accessibility, unmanned aerial systems are considered particularly compatible with cable-stayed bridges. On its database of such bridges, the Federal Highway Administration profiles 28 cable-stayed spans on the continent, but other research indicates the total could be closer to 60. In Europe, where this type of design was implemented earlier than in the United States, unofficial research indicates more than 1,600 cable-stayed bridges. Thus, the use of UAS to assist bridge inspectors could have global implications.

Last year that theory was put to the test on an inspection of the William H Natcher Bridge, a cable-stayed bridge over the Ohio River between Kentucky and Indiana. The bridge owner, the Kentucky Transportation Cabinet was enthusiastic about the UAS trial; as a cable-stayed structure with non-fracture critical cables, it provided the ideal test-bed study, with relatively unimpeded access to all cables and low traffic volumes. The structure was also subject to a hands-on inspection at the same time.

This pilot study validated the advantages of UAS and confirmed the results of earlier inspections, allowing KYTC to monitor the progress of previously identified defects. It enabled even the smallest characteristics to be identified, including a newly discovered 50mm crack in the superficial but protective high-density polyethylene sheath on one cable, and it reduced the time the work would have taken if rope access methods only had been used.

At the same time, the trial helped identify some limitations to UAS — at least in its current state. Those limitations may be ameliorated or eliminated as the technology evolves, but for now, all those involved in bridge inspection, including owners and contractors, should understand the limitations as well as the advantages.

The 2017 routine and fracture critical inspection project was part of KYTC’s biennial inspections of the Ohio River Bridges; it included an inspection of the Natcher Bridge, which was opened in 2002 and which with its approaches is 1,373m long.

As the project was planned with KYTC, the team jointly envisioned a hybrid for testing purposes as this was the first time the cables were scanned at length. Initially, the inspection would be conducted by traditional methods and rope access, with half the cables scanned by UAS requiring hands-on validation. However, when inspectors rappelled down the cables, they were buffeted by strong winds in excess of 65km/h.

These conditions prompted the team to modify their approach. As the high winds made the rope access team’s time less efficient, the team chose to inspect all the non-fracture cables between the anchorages on the bridge with UAS at a later date, when windspeed had decreased; the unit used for the flight had a maximum wind resistance of 10m/s. Rope access inspectors would inspect all cable anchorages along the tower faces and at the deck, accessible in the winds, and use the 24 cable lengths, inspected hands-on via rope previously on calmer days, as visual validation of UAS results. This approach resulted in an even more extensive trial than originally planned.

Any bridge operators or contractors considering a UAS approach or a combined UAS-rope inspection plan, will have to take into account quite a few federal, state and even local regulations. For rope access activities, inspectors should be certified by the Society of Professional Rope Access Technicians or Industrial Rope Access Trade Association. Also, the National Highway Institute certification for safety inspection of in-service bridges as well as fracture-critical inspection techniques for steel bridges may be required by local agencies.

Because an unmanned aerial system cannot operate over non-participants or directly above live traffic, a lane of traffic had to be closed for much of the procedure. However, this restriction permitted underbridge inspection vehicles to be used for the floor system inspections at the same time. Flights through controlled airspace would require coordination with the Federal Aviation Administration, but the Natcher Bridge is in uncontrolled Class G airspace, so a pilot with a Part 107 licence can fly it without the need for additional authorisations or waivers. In this case, therefore, it was an unusually light regulatory component.

The next step was selecting the most appropriate airframe, or drone, for this project. Over recent years, some engineering firms have made a concerted effort to build robust UAS fleets. For example, Michael Baker International has amassed approximately 40 airframes and 38 pilots. Developing such a capability is not a step to be taken lightly, as the costs can be significant. An entry-level airframe costs about US$1,500, but with add-ons such as sophisticated sensors, the price can mount quickly. As a point of comparison, one of the most functional sensors is a Panasonic Lumix 30x travel zoom unit paired with a Flir Tau 640 thermographic sensor. With prices proportional to the capability of the equipment, unit costs can exceed US$55,000 in many cases.

For this project, the DJI Inspire 2 airframe was used. It features a dual-battery system that can extend flight time as well as self-heating technology that allows it to function in lower temperatures. The Inspire 2 is a favorite of filmmakers because it can accommodate a variety of video formats. The team decided that video, rather than still photography, would provide the best record for both the current project and as a benchmark for future inspections.

In all, the Inspire 2 scanned 96 cables ranging in length from 137m to 182m. While an experienced UAS pilot flew along the cables, a sensor operator and visual observer – also a licensed UAS pilot – independently operated the camera to ensure quality scans of the cables for deficiencies. Experienced inspectors then reviewed the footage to ensure broad consensus about the data collected and validated findings by the UAS team. The results of the UAS inspection were encouraging. The UAS scanned each cable in an average of about eight minutes after rhythm was developed, compared to the 45 minutes the inspection by rope access alone would have required.

The UAS scans took two days to complete with another day’s time devoted to video processing and review; however, the review was conducted in parallel in case rope inspectors had to investigate further. Had the inspection used rope-access exclusively, it was estimated it would have taken five to seven days with the aid of motors to also ascend the cables; the 96 tower connections took approximately two to three days to rappel. Not only do these results strongly suggest that UAS can improve efficiency and reduce costs, but the positive impact on worker safety and reduced exposure, where applicable, is a valuable benefit.

These advantages were acknowledged by Evan Dick, KYTC’s transportation engineer specialist, who worked closely with the team on the project. “Through our limited use of UAS on bridge inspections, we have found value in both added safety and cost savings,” Dick said. “During the inspection of the cable stays on the Natcher Bridge, scanning the stays with UAS decreased the time needed for an inspector to be on rope by up to 75%. At this time, UAS can be a valuable tool when used in the right situations.”

As it can cut days from the inspection, another plus of using the UAS is that it can reduce disruption to traffic on the bridge. Furthermore, the inspection had no impact on river vessels.

The Inspire 2 accurately captured all the defects previously noted at anchorage locations during rope inspections. Widespread cracking was identified in the HDPE sheaths at anchorage locations ranging from hairline to 20mm cracks detectable from 3m to 12m away by UAS. Most impressive of all, the airframe detected a new 5mm to 1.5mm wide by 215mm long crack mid-length in the cable casing. A follow-up hands-on inspection by rope access indicated the casing was leaking grease rather than rust.

There was also a slight bulge in the casing possibly indicating cable relaxation. The finding was not yet a major safety issue, but its detection enabled KYTC to note the flaw for future observation and repair. This experience suggests that UAS can help identify locations for rope access crews to focus their efforts, yet another valuable time saver.

The UAS flight technique entailed a flight path along the length of the cable, with the intent to remain roughly 3-5m away from the cable. The pilot worked to ensure the flight path remained consistent along the length, while the sensor operator fine-tuned the camera angle and focus to capture any cable defects. If either the pilot or sensor operator viewed a defect during flight, the UAS stopped to take additional footage of the defect.

Otherwise, the data was viewed by the bridge inspection team onsite and, if an area needed further investigation, the pilot took the UAS back to that portion of the previously recorded cable and more video and/or pictures were taken.

While the Natcher Bridge pilot study largely confirmed the advantages of using unmanned aerial systems, it also pointed out a few limitations. Past experience suggests its value may be restricted on bridges where access is not as open as that found on cable-stayed bridges. If, for example, an inspection is performed between girders, beneath the structure or in a more nuanced area – gusset plate nodes on trusses for example – the pilot risks losing the GPS signal and having to rely on manual navigation. In such situations, particularly where heavy wind makes the airframe unstable, the pilot may not be able to manoeuvre the unit into difficult-to-access spots. If the airframe crashes, not only is it lost but so is its valuable video cargo, to say nothing of the time and financial resources already invested. A formal federal investigation may also be mounted if there is loss of property or injury.

One of the most important lessons learned about the use of unmanned aerial systems was provided by Jeff Sams, a technical specialist in Michael Baker’s Louisville office who is also an inspector and former chief bridge inspector for KYTC. Sams assisted with the inspection preparation by designing test flights for the UAS team to validate scanning tolerances. “You have to double check what the machine sees versus what is acceptable for the inspection threshold,” Sams advises. “If your airframe can’t see a 1.5mm wide crack, that doesn’t mean it’s not there.”

On the test flights, Sams made sure that the unit could detect flaws to the exact tolerances KYTC specified and he conducted test flights on both the sunny and shady sides of the bridge to establish appropriate benchmarks in each case. It was noted the UAS was incapable of observing the internal strands at HDPE crack locations at anchorages where inspectors had to physically manipulate the HDPE to give them sufficient light to examine the structural strands.

Unmanned aerial systems represent a fantastic new tool in the inspection tool box, but should not be viewed as a panacea or a replacement for hands-on inspection. They can enhance and provide focus for traditional or rope access inspections, helping bridge owners save financial resources, protect their human resources and maintain a permanent, researchable record of inspections for long-term structural health monitoring.

Continued investigation and data verification of UAS capabilities is planned in the near future. A prestressed concrete girder bridge was recently scanned and early results indicate hairline cracking is detectable, but further processing is required. The team also intends to return to the Daniel Carter Beard Bridge in Cincinnati where a full rope access inspection of the steel tied arch was carried out two years ago. During the next routine and fracture-critical inspection, a UAS validation study will be carried out on the hanger cables of this bridge.

John Zuleger is technical manager and rope access programme manager and Alicia McConnell is a UAS remote pilot and transportation engineer; both work for Michael Baker International.