Using drones to find broken bridges – before it's too late

Using drones to find broken bridges – before it's too late

October 12, 2017
By Matt Windsor
With a $500,000 grant from the National Science Foundation, UAB engineers explore an innovative drone strategy to address America's crumbling infrastructure.

America has a big bridge problem.

There are more than a million bridges in the United States, and most were built during the great highway construction boom of the 1950s. They were designed for the relatively light panel trucks of the Eisenhower era, rather than the massive rigs of today. Meanwhile, the frequency of trips has surged. In 1950, trucks hauled 200 billion ton-miles of intercity commercial freight; by 2002, that number increased to 1.44 trillion ton-miles, or more than 600 percent.

mix bridge mississippiInspecting massive bridges like the Mississippi River span outside Memphis is a herculean task. “Cross [that bridge], and you’ll understand,” says UAB researcher Nasim Uddin. “How can you expect a human to reach all of it?”

These bridges had an average design life of 50 years, meaning they’re all on borrowed time. According to the U.S. Department of Transportation’s official list, there were 16,098 “structurally deficient” bridges in Alabama as of 2016: 15,935 state- or county-owned, and 163 Federal bridges. [A bridge is “structurally deficient” if it receives a condition rating of 4 or less (on a 10 point scale) on its deck, superstructures, substructures or culvert and retaining walls.]

“Those should all be decommissioned now,” says Nasim Uddin, Ph.D., professor of civil engineering in the UAB School of Engineering. But tearing them down and building new bridges would cost trillions of dollars. So the Department of Transportation is focused on “how to maintain them — how to live with them,” Uddin says.

Regular inspections are crucial to spot warning signs before disaster strikes, as it did with the collapse of the I–35 bridge in Minneapolis, which killed 13 and injured 145 in 2007. But doing that job right would require hiring a massive amount of new inspectors, and learning all the signs of a failing bridge can take decades of experience.

Besides, inspectors often can’t reach all areas of a bridge. “Cross the Mississippi River bridge outside Memphis, and you’ll understand,” Uddin says. “How can you expect a human to reach all of it?”

With a nearly $500,000 “breakthrough” grant from the National Science Foundation, Uddin is testing three interrelated ideas.

 mix drone sensors 700

1. Install sensors on bridges to provide an early warning sign of problems.

UAB researchers have already tested similar technologies during earlier studies. “This is the culmination of a number of projects we’ve been doing with the NSF,” Uddin says. There are two issues: power and transmission. How do you keep the sensors running, and how do you collect the data they have recorded?

“The sensors need power,” Uddin says. “A battery would die within a few days. If you have access to continuous electricity from a nearby city, that’s great. Unfortunately, many bridges have no power, being far away from cities.” The cost of installing electricity and data transmitters to hundreds of thousands of bridges is cost-prohibitive.

That leads to step 2: enter the drones.

mix drone drones 700 

2. Fly drones out to the sensors on a regular schedule to collect data.

The answer, says Uddin, comes in the form of compact “mosquito” drones. “We are going to have drones, flying on autopilot, hover next to the sensor and charge it up” through induction technology, similar to the wireless charging devices now available for smartphones, he says. At the same time, the drone will collect sensor data that can be uploaded to the cloud or brought back to engineers at its base. A cage around the drone will protect the bridge, and any passing humans, from interacting with the machine’s propellers.

 mix drone charging 700

Drones themselves suffer from limited battery capacity, however. So where will their power come from? Uddin envisions a network of charging pads, spaced along the drone’s route between its home base and the bridges. Another idea: equipping drones with special hooks that would let them recharge from power lines on the fly.

If the mosquito drones report a problem, a truck carrying a larger imaging drone could head to the bridge for a drive-by inspection. The drone would use sensors such as LIDAR and thermal imaging cameras that “can look through the concrete, like an X-ray, to see if the rebar inside has developed any cracks. They can see the hidden dangers,” Uddin says. Meanwhile, the truck itself would carry sensors to measure bridge stability as a second check.

 mix drone scan data 700

3. Automatically scan the data to alert human engineers when a bridge is failing.

The third aspect of the UAB approach is the development of damage detection algorithms that can automatically interpret all this sensor data. “No one can afford to have engineers looking at the data all the time,” Uddin says. His team is studying whether short data bursts from the sensors — rather than a continuous stream of information — can provide the vital information engineers need to decide whether a bridge has become unsafe.

The UAB team isn’t working alone. Teams from Queen’s University Belfast, in the United Kingdom, and University College Dublin, in Dublin, Ireland, are pursuing related projects. Each university has received around $500,000 grant from its respective government funding agencies, leveraging the NSF project. “Our students and faculty will interact with each other” — with the UAB team visiting their counterparts for research, for example, Uddin says. University College Dublin and Queen’s University Belfast have indoor labs for bridge testing and instrumented bridge sites that are important for the success of the project, he notes. However, UAB faculty and students could do early testing of their drone system here, without the dangers of field testing, Uddin adds.

“This is a unique experience that we will be able to offer engineering students at UAB,” Uddin says. “They will take part in projects with an opportunity to make a tremendous difference.”

Related story: Two other UAB engineering researchers recently won significant national awards to fund innovative highway safety studies