By Jenny Jones
Tufts University engineers develop a concept for using drones and vibration sensors to automate the bridge-inspection process.
Two engineering research teams from Tufts University are working together to develop a system that uses wireless vibration sensors, cameras, and drones in a coordinated effort to monitor the structural stability of bridges in real time. Wikimedia Commons/U.S. Army Corps of Engineers
October 7, 2014—With separate funding from the National Science Foundation (NSF), one engineering team at Tufts University has been developing damage-detection algorithms for use in automated bridge inspections and another has been developing technologies for autonomous flying robots, known colloquially as drones. Now the two teams are working together to develop a conceptual design for a system that uses wireless vibration sensors, cameras, and drones in a coordinated effort to monitor the structural stability of bridges in real time.
The Federal Highway Administration's National Bridge Inspection Standards require bridge owners, typically state departments of transportation, to perform safety inspections of bridges measuring 20 ft long or more every two years. These inspections traditionally involve shutting down a portion of a bridge and using a boom truck or other apparatus to visually examine the structure. The time-consuming process is disruptive, costly, accident prone, and somewhat subjective.
"There is a huge need for better bridge-inspection techniques," says Babak Moaveni, Ph.D., A.M.ASCE, an associate professor in the university's department of civil and environmental engineering, who is developing the combined system with Usman Khan, Ph.D., an assistant professor in the university's department of electrical and computer engineering.
ASCE's infrastructure assessment reports have long indicated that the nation's bridges are in dire condition, its 2013 Report Card on America's Infrastructure giving bridges a grade of C+. And a July report from the Obama administration echoes the Society's findings. The federal report states that one in four of the nation's bridges is in need of significant repair and unable to handle current traffic loads. Frequent inspections are critical to ensure that structural deficiencies are identified, monitored, and addressed before they lead to catastrophic events.
In Moaveni and Khan's system, vibration sensors will be used to identify those deficiencies. Equipped with radio frequency identification (RFID) chips and computer processing units, the so-called smart sensors will be attached to a bridge and programmed to collect vibration data from the structure at certain intervals. The sensors will then process that data so that only pertinent findings are stored on the RFID tags. Changes in vibration signals could indicate that the bridge has been damaged, says Moaveni, who has received a grant from the National Science Foundation (NSF) for research to further the development of the wireless bridge inspection system.
When it comes time to inspect the bridge, drones equipped with computers will hover close to the sensors to extract the data from the RFID chips. If the data indicate possible damage at a particular location, the drones' onboard cameras will take pictures of that area. The drones will wirelessly transmit the data and images to a laptop or central computer, where a color-coded system of green, yellow, and red will alert inspectors of the bridge's status. "Green means there's no need to inspect, yellow means there could possibly be something, and red means for sure there is something wrong with the structure," Moaveni says. "Instead of inspecting the bridge every two years, the inspectors will go and visit the bridge, for example, only when yellow or red is transmitted to them."
The drones will also be capable of reprogramming the sensors. For instance, if the bridge owner knows that the bridge will be under a particularly high amount of stress at a certain time, the drones could be sent to the structure to tell the sensors to collect data more frequently during that period. The drones may also start this process automatically based on weather or traffic alerts monitored by the drones. Or if the amount of data that the sensors have collected is insufficient, the drone can tell the sensors to "wake up" to collect more data, Moaveni says. Once the drones are finished collecting data or performing other tasks at the bridge site, they will either fly back or be taken back to their home location, where they will be recharged for the next operation.
While the sensors will require power to process the data and add data to the RFID tags, power will not be required to transmit that data to the drones. As a result, the sensors will be able to run from 5 to 15 years on batteries that are slightly more powerful than standard cell phone batteries, Khan says. Khan obtained his NSF grant to conduct research on the drones, which could inform the development of the bridge-inspection system. "Most of the battery [power] in these sensors is consumed when they are transmitting data from one location to another," Khan explains. "We have eliminated the need for this long-distance communication because the robot is actually going to the sensor...and using electromagnetic fields to read and write the data on the RFID tags without requiring any power from these tags."
Developing the concept into a working system presents many challenges. Researchers must figure out how to coordinate the drones so that no two drones collect data from the same sensors and so that the drones collect the data in the most efficient manner. They must also figure out how to keep the drones on course when they travel under the bridges—where the global positioning system signal they need for navigation is often weak or not available—and under varying weather conditions, Khan says. "One way we can solve this problem is to send fancy and very expensive robots, but that defeats the purpose if the solution is expensive," he says. "We want to send cheap robots for the solution to remain scalable and economically viable."
The researchers are currently exploring funding options to solve the system challenges and take it beyond the concept phase to a prototype that can be tested on an actual bridge. Moaveni says that the ultimate goal is to create a working system that will reduce the number of hours engineers must spend visually inspecting bridges, leading to significant cost savings and more reliable data. "We're not proposing to completely replace the current approach to inspecting bridges, but we are trying to create tools to complement and improve it," he says.