by The deteriorating state of our built environment, reflected in recent building collapses and structural failures of roads and bridges, is posing a growing concern. The maintenance efforts are unable to keep up with the rate at which our infrastructure is aging and deteriorating. The sheer scale of the problem makes it impractical to manually inspect every crack and defect to identify signs of potential failure amidst the normal wear and tear. To tackle this issue, researchers at Drexel University’s College of Engineering have developed a system that utilizes artificial intelligence (AI) to assist robotic assistants in the inspection process.
By combining computer vision and a deep-learning algorithm, the researchers have devised a multi-scale system that can efficiently identify problem areas and conduct inspections using autonomous robots. The system, outlined in a study published in Automation in Construction, starts with visual inspection technologies that have previously been used for damage assessment. However, the addition of AI enables a more precise and automated identification of cracks and other potential issues. The system then directs laser scans to create a digital twin computer model, which aids in assessing and monitoring the damage.
The proposed system offers a significant reduction in the overall inspection workload, allowing inspectors to focus on critical areas that require immediate attention. In the study, the team likens cracks to medical symptoms that should be screened in their early stages. Early and accurate detection of cracks is crucial for timely diagnosis, maintenance, and repair efforts, preventing further deterioration and potential hazards.
Given the immense backlog of repairs needed for the nation’s infrastructure, setting up a triage system is of utmost importance. The Bipartisan Infrastructure Law estimated a backlog of $786 billion in repairs to roads and bridges prior to its implementation. However, the repair process is further complicated by a shortage of skilled infrastructure workers, including inspectors and repair specialists.
According to Arvin Ebrahimkhanlou, an assistant professor at Drexel’s College of Engineering, civil infrastructures, including large-scale structures and bridges, often have defects that are small-scale in nature. To address this, the researchers propose a multi-scale robotic approach that combines computer vision for efficient pre-screening of problem areas and precise robotic scanning of defects using nondestructive, laser-based scans.
The system moves away from subjective human interpretation by adopting a high-resolution stereo-depth camera feed that is processed by a deep-learning program called a convolutional neural network. These AI programs, known for their ability to detect patterns and anomalies in large quantities of data, have been used for applications such as facial recognition, drug development, and deepfake detection.
The proposed AI-guided system has the potential to revolutionize the inspection process for buildings, roads, and bridges. By combining computer vision, deep learning, and robotics, inspectors can efficiently identify and assess defects, helping prioritize repairs and prevent potential structural failures. As our infrastructure continues to age, innovative solutions like these are crucial to ensure its safety and longevity.
*Note:
1. Source: Coherent Market Insights, Public sources, Desk research
2. We have leveraged AI tools to mine information and compile it
