Scientists Develop Infrastructure Inspection Robot


Researchers at Drexel University are currently studying ways to use robotic assistants in identifying and inspecting problem areas in infrastructure such as buildings, roads and bridges.

According to a release from the university, a new multi-scale system developed by the team uses computer vision and machine learning programs to identify cracks in concrete and direct robotic scanning, modeling and monitoring.

About the System

The university says that the environment is aging faster than can be maintained. Recent building collapses and structural failures of roads and bridges reportedly indicate a problem that could get worse, according to experts, because it’s just not possible to inspect every issue and parse dangerous signs of failure from normal wear and tear.

To combat this, the team at Drexel began working on augmenting visual inspection technologies with a new machine learning approach to find any cracking and directs a series of laser scans of the regions to create a digital computer model to assess and monitor the damage.

According to the team, the system showcases a strategy that could help reduce the overall inspection workload and allow for focused consideration and care to mitigate structural failures.

“Cracks can be regarded as a patient’s medical symptoms that should be screened in the early stages,” the authors, Dr. Arvin Ebrahimkhanlou, an assistant professor, and Ali Ghadimzadeh Alamdari, a research assistant, both in Drexel’s College of Engineering, wrote.

“Consequently, early and accurate detection and measurement of cracks are essential for timely diagnosis, maintenance, and repair efforts, preventing further deterioration and mitigating potential hazards.”

The team states that many of the nation’s buildings, bridges, tunnels and dams are still in danger of failure, urging that the response should be to set up a triage system.

Before the bipartisan infrastructure law, the American Society of Civil Engineers reportedly estimated a backlog of $786 billion in repairs to roads and bridges. The release states that a shortage of skilled infrastructure workers, including inspectors and those who could repair aging structures, has added to this issue.

“Civil infrastructures include large-scale structures and bridges, but their defects are often small in scale,” Ebrahimkhanlou said. “We believe taking a multi-scale robotic approach will enable efficient pre-screening of problem areas via computer vision and precise robotic scanning of defects using nondestructive, laser-based scans.”

Instead of a physical measurement by human eyes, the system reportedly uses a high-resolution stereo-depth camera feed of the structure into a deep-learning program called a convolutional neural network.

These programs, reportedly also used for facial recognition, drug development and deepfake detection, have gained attention for their ability to find patterns and discrepancies in large volumes of data. Training the algorithms on datasets of concrete structure images then reportedly turns them into “crack-spotters.”

“The neural network has been trained on a dataset of sample cracks, and it can identify crack-like patterns in the images that the robotic system collects from the surface of a concrete structure. We call regions containing such patterns, regions of interest,” said Ebrahimkhanlou, who leads the research on robotic and artificial-intelligence based assessment of infrastructure, mechanical and aerospace structures in Drexel’s Department of Civil, Architectural, and Environmental Engineering.

Once the cracked or damaged area is identified, the program reportedly directs a robotic arm to scan over it with a laser line scanner, creating a three-dimensional image of the damaged area. Concurrently, a Light Detection and Ranging (LiDAR) camera scans the structure around the crack. 

Stitching both plots together reportedly helps create a digital model of the area to show the width and dimensions of the crack and enables tracking changes between inspections.

“Tracking crack growth is one of the advantages of producing a digital twin model,” Alamdari said. “In addition, it allows bridge owners to have a better understanding of the condition of their bridge, and plan maintenance and repair.”

The release states that the team ran the system in the lab on a concrete slab with several different cracks and deterioration. In the testing, the system was reportedly sensitive enough to find and size up fissures that were less than a hundredth of a millimeter wide.

According to the team, while human inspectors would still make the final call on when and how to repair the damages, the robotic assistants could help reduce the workload.

Additionally, an automated inspection process, the team says, could reduce oversights and subjective judgement errors that can occur when overworked human inspectors take the first look.

“This approach significantly reduces unnecessary data collection from areas that are in good structural condition while still providing comprehensive and reliable data necessary for condition assessment,” the team wrote.

The researchers added that they want to incorporate the multi-scale monitoring system as part of a larger autonomous monitoring framework with drones and other autonomous vehicles.

As an example, the team pointed out one proposed by the U.S. Federal Highway Administration’s Nondestructive Evaluation Laboratory, which would utilize tools and sensing technologies to autonomously monitor and repair infrastructure.

“Moving forward, we aim to integrate this work with an unmanned ground vehicle, enhancing the system's ability to autonomously detect, analyze, and monitor cracks,” Alamdari said.

“The goal is to create a more comprehensive, intelligent and efficient system for maintaining structural integrity across various types of infrastructure. Additionally, real-world testing and collaboration with industry and regulatory bodies will be critical for practical application and continuous improvement of the technology.”

The team’s research was published in the journal Automation in Construction.

Other Robotic Inspections

In January, report from advanced robotics and enterprise software company Gecko Robotics and climate technology company Rho Impact detailed how robots and artificial intelligence could improve their data collection to help maintain crumbling infrastructure and bring about a zero-carbon economy.

The report was published in the World Economic Forum and detailed how despite AI’s potential to stop the negative effects of climate change, the issue of data collection had often been overlooked.

According to the report, about 90% of private and public sector CEOs have stated that AI could be an essential tool to fight against climate change; however, 75% have stated that they didn’t have much trust in the data that is collected.

Gecko explained that AI models are only as good as the data they have trained with, relying on “robust and granular data sets” to find patterns and trends to allow the models to learn and develop predictive capability.

The report stated that AI's impact on people’s daily lives had shown the potential to aid in emissions reduction in critical infrastructure.

Data-driven insights for decision-making are reportedly often relied on in daily life. As an example, a November 2023 Boston Consulting Group report detailed how AI-optimized transportation has aided drivers in finding more efficient routes, subsequently cutting emissions.

Additionally, the success of AI-driven thermostat adjustments had reportedly helped save a total of 113 billion kilowatts of energy, about double Portugal's annual electricity use between 2011 and 2022.

Gecko said that the shift towards data-informed decision-making can produce a strong base for the broader application of AI, with an emphasis on the role of accurate data in achieving net-zero emissions in important infrastructure across sectors.

According to the report, advances in robotics, sensors, drones and other data-collection technologies allowed for the collection and analysis of data on the built world in ways that just five years ago were not attainable.

These innovations, the report added, can offer new access to ground truth data gathered from the physical world.

Technologies such as robots armed with sensors like ultrasonics that crawl the surface of infrastructure, producing digital twins with high-fidelity data layers about asset health, have aided in addressing challenges that had previously been overlooked.

This new shift could reportedly be the difference between seamless operations and catastrophic failures at places like oil and gas refineries, power plants and manufacturing facilities.


Tagged categories: Asia Pacific; Bridges; Bridges; Colleges and Universities; concrete; Concrete defects; Construction; Digital tools; EMEA (Europe, Middle East and Africa); Health & Safety; Health and safety; Infrastructure; Infrastructure; Inspection; Inspection equipment; Latin America; North America; Program/Project Management; Research; Research and development; Roads/Highways; Robotics; Safety; Technology; Tools; Tools & Equipment; Z-Continents

Join the Conversation:

Sign in to our community to add your comments.