BRIDGE RESCUE: (Bridge) inspector gadget

FHWA, Rutgers roll out bridge-deck assessing robot to collect quality data for LTBP program

Jeff Zagoudis / May 08, 2014
RABIT Bridge Deck Assessment Tool
RABIT Bridge Deck Assessment Tool

Pulling a rabbit out of a hat is a sure-fire way for a magician to impress an audience. The researchers at the Federal Highway Administration (FHWA) and Rutgers University just pulled out a different kind of rabbit, and bridge owners across the nation are certainly taking notice.

In this case, however, the RABIT Bridge Deck Assessment Tool is a robot. It represents the future of bridge-deck inspection, combining a wide array of current technologies into one platform.

The robot was generated from another forward-thinking endeavor—FHWA’s Long-Term Bridge Performance (LTBP) program. Spawned by Congressional mandate in 2008 (from 2005’s SAFETEA-LU), the LTBP program is a research program examining the performance of the nation’s highway bridges over an extended period.

Birth of a bridge program
Once the program was established, FHWA spoke with the various stakeholders and partners of the LTBP program to determine where they should focus their research efforts. The No. 1 priority, according to Hamid Ghasemi, Ph.D., team leader of the Infrastructure Management Team in FHWA’s Office of Infrastructure Research and Development, was the performance of the bridge deck—specifically concrete bridge decks.

After priorities were established, FHWA looked at the big picture—all 600,000 bridges in the national bridge inventory—and identified the most common types, seven in all. “Then we asked what information do we need to have a good understanding of the performance of concrete bridge decks,” Ghasemi said. “What tools do we need and how do we need to go about this?”

Finding the tools was not the problem. A wide array of technologies suited for bridge-deck inspection were already in use, and FHWA wanted to focus particularly on nondestructive evaluation (NDE) methods. NDE tools and methods allow researchers to gather information about a structure without permanently changing or damaging the bridge.  

Dreaming big
The need for an automated solution hit home for Ghasemi and his team while he was participating in the initial inspection of an LTBP program pilot bridge in Haymarket, Va. FHWA wanted to put the various pieces of technology through their paces and make sure they had a baseline for data collection. During the inspection, Ghasemi looked around and realized there were close to 30 people on the bridge. “Imagine doing that for the large sample of bridges all across the country,” he said. “There was no way, with the resources we have, that we could do that.”

The other part of it was ensuring that data-collection procedures and methods remained perfectly consistent for every bridge inspection—something difficult to guarantee, especially with such a large number of people working simultaneously.

So, in 2011, Ghasemi, Firas I Sheikh Ibrahim, Ph.D., P.E., and the Infrastructure Management Team, under the direction of Jorge Pagan-Ortiz—director of the Office of Infrastructure Research and Development—began working on what would become the RABIT Bridge Deck Assessment Tool. The idea was firmly in place, but actually putting the platform together would be the most difficult task. For help, FHWA turned to the Center for Advanced Infrastructure and Transportation (CAIT) at Rutgers University. Led by Drs. Ali Maher and Nenad Gucunski, CAIT brought experts in automation, data fusion and other key technical areas to the team.

“Data integration, synchronization, visualization and consistent interpretation were the key challenges,” Ghasemi told ROADS & BRIDGES. “We had to consider whether we could make sense out of each technology and at the end of the day, can you fuse the data together?”

The first step was determining which technologies would be part of the automated platform. The idea was to have the robot collect as much meaningful data as possible in one pass, so the development team wanted to include a full suite of NDE technologies. They ultimately chose to include:

Impact echo;
Ultrasonic surface wave testing;
Ground-penetrating radar;
Electrical resistivity; and
High-resolution digital cameras.

GPS technology and wheel decoders also were incorporated into the platform, which proved to be the secret to its automation—and the greatest challenge for the development team to master. The idea was for the machine to be guided along a precise path using GPS coordinates. To improve safety for the robot and its handlers, the team also installed a series of laser sensors, allowing it to detect and avoid obstacles.

Out in the field
The way Ghasemi explains it, the RABIT Bridge Deck Assessment Tool inspecting a bridge deck is no different than a person going to see a doctor. “When you go to the doctor, it’s not a one-size-fits-all thing; they’re not going to give you an aspirin to solve everything,” he said. “They’ll look you over, maybe give you an EKG or an X-ray to see what’s wrong with you. And this robot does the same thing in one device.”

Initial testing was done in a parking lot on the Rutgers - New Brunswick campus. The team also utilized several small bridges in New Jersey to conduct early validation testing.

When the time came for field deployment, FHWA selected two of the seven pilot bridges tabbed for the LTBP program. They went back to Virginia for the first test in October 2012, choosing a three-span bridge in Loudoun County along westbound S.R. 7. Average daily traffic on that particular bridge was approximately 24,500 vehicles.

For the second field test a month later, the team stayed in Rutgers’ backyard—specifically Gloucester County, N.J. The bridge they inspected there was a two-lane, single-span structure with an average daily traffic count of 18,000 vehicles.

In both cases, the LTBP team had already completed a visual inspection of the bridge deck with handheld NDE technologies, and the robot was sent in to validate the data. Rather than the large groups FHWA had deployed earlier, only three researchers accompanied the robot to the inspection sites, remaining inside a control van while the machine did its work. According to Rob Zobel, Ph.D., P.E., development and technical engineer for the LTBP program, programming is fairly simple: “Define the extent of what you want the RABIT to look at, push a button and it’ll go,” he told ROADS & BRIDGES.

And go it did—condition assessment of each deck was completed in a day’s work (roughly four hours), significantly faster than using conventional handheld NDE tools.

Landmark case
The RABIT Bridge Deck Assessment Tool got its biggest test to date when the Infrastructure Management Team got a call from the Eastern Federal Lands Highway Division (EFLHD) and the National Park Service in February 2013 with a request for its services on the Arlington Memorial Bridge in Washington, D.C. Eighty years old at the time, the bridge was scheduled for rehabilitation, so EFLHD wanted to get a sense for the condition of the aging giant’s bridge deck. Specifically, the agency was concerned about degradation, corrosion, delamination of the bridge deck and debonding in the asphalt overlay, which ranges from 2 to 4 in. thick.

As a primary connector into the D.C. area (and a part of the National Register of Historic Places), the 2,138-ft-long, six-lane bridge sees a heavy amount of daily traffic, meaning that shutting it down completely was not an option. FHWA deployed the robot to analyze three lanes. Moving at a rate of roughly 4,000 sq ft per hour, the robot was able to complete the entire inspection in just 15 hours.  

According to Zobel, doing the same analysis the old-fashioned way could have taken three to four weeks.

Best of all, the data collected by the RABIT Bridge Deck Assessment Tool on the Arlington Memorial Bridge confirmed the findings from core samples. “As a result of this, EFL got a better understanding of the extent of deterioration in the bridge deck,” which helped convince the National Park Service to replace the bridge deck, said Zobel.

The next steps
Both Ghasemi and Zobel confirmed that all of the data collected by the RABIT Bridge Deck Assessment Tool so far has been validated, allowing them to think about the future of the automated bridge tool.

The team has heard concerns that tools such as this will eventually take the place of visual inspection, but Ghasemi quickly and adamantly disputed that misconception. “Don’t think you can ever replace the visual inspection, no matter what technology we have,” he told ROADS & BRIDGES.

At present, FHWA plans to deploy the RABIT Bridge Deck Assessment Tool on LTBP bridges in the mid-Atlantic region for further testing and then take it on a roadshow soon to state departments of transportation and other agencies across the U.S. to illustrate what the machine can do. “We would like to do demonstrations on bridges in different regions,” Ghasemi said, to see how the machine behaves in different conditions.

FHWA currently has two prototypes and in collaboration with Rutgers plans to build more as the word gets out and the demand for the RABIT Bridge Deck Assessment Tool rises. The Infrastructure Management Team also has begun discussing other technologies that could be used in a similar tool to assess other parts of a bridge. R&B


SIDEBAR

Tools of the trade

The RABIT Bridge Deck Assessment Tool combines a number of nondestructive evaluation (NDE) technologies to provide one comprehensive diagnostic of a bridge deck after just one pass. According to FHWA’s Rob Zobel, most of these methods can be used quantitatively and/or qualitatively as needed:

Ultrasonic surface wave (USW) testing is used to create a profile of the modulus of elasticity of the concrete bridge deck. While an exact number can be beneficial, the greater benefit is being able to compare measurements at different locations, according to Zobel.

Impact echo (IE) detects delaminations (horizontal cracks) inside the concrete using a small impact device to create a stress wave that travels through the concrete, reverberating and creating small, measurable displacements at the surface. These displacements can be used to determine the presence of and depth to a hidden defect. The robot uses IE to create a surface contour map of the bridge deck indicating location and severity of any defects.

Ground-penetrating radar (GPR) looks for deterioration by bouncing electromagnetic waves off embedded reinforcement. Attenuation of the reflected waves indicates a defect in the concrete and/or reinforcement corrosion.

Electrical resistivity is another corrosion test that involves sending electric current through the concrete between a series of probes, measuring the voltage drop and calculating resistivity of the concrete. Greater resistivity equals denser concrete and greater difficulty for corrosive chemicals to migrate through the concrete.

As the RABIT Bridge Deck Assessment Tool moves across the bridge deck, it transmits all of the gathered data back to the van, where it is analyzed and correlated into a near real-time “map,” giving researchers a 3-D image of what is happening inside the concrete deck.

About the Author

Related Articles

It is easy to forget in our coverage of bridges and bridge building that some of the most used and useful structures don't "bridge" roadways at all.…
May 07, 2021
The U.S. faces an unprecedented challenge in maintaining and rehabilitating its aging transportation infrastructure. Agencies in every state are…
May 05, 2021
A study recommended using a remotely operated, sensor-equipped unmanned surface vessel (USV) to support the Michigan DOT’s bridge inspections.
A study recommended using a remotely operated, sensor-equipped unmanned surface vessel (USV) to support the Michigan DOT’s bridge inspections. All photos: Michigan DOT
Michigan wants all of its bridges to be on solid footing. In 2019, the Michigan Department of Transportation (MDOT) commissioned a study to look at…
May 05, 2021
The Third Avenue bridge connecting downtown Minneapolis to the city’s northeast side needed rehabilitation. Everyone agreed on that—but the bridge…
May 05, 2021
expand_less