Development of AI- and Robotics-Assisted Automated Pavement-Crack-Evaluation System

Abstract

Crack inspection is important to monitor the structural health of pavement structures and make the repair process easier. Currently, pavement crack inspection is conducted manually, which is inefficient and costly at the same time. To solve the problem, this work has developed a robotic system for automated data collection and analysis in real-time. The robotic system navigates the pavement and collects visual images from the surface. A deep-learning-based semantic segmentation framework named RCDNet was proposed. The RCDNet was implemented on the onboard computer of the robot to identify cracks from the visual images. The encoder-decoder architecture was utilized as the base framework of the proposed RCDNet. The RCDNet comprises a dual-channel encoder and a decoder module. The encoder and decoder parts contain a context-embedded channel attention (CECA) module and a global attention module (GAM), respectively. Simulation results show that the deep learning model obtained 96.29% accuracy for predicting the images. The proposed robotic system was tested in both indoor and outdoor environments. The robot was observed to complete the inspection of a 3 m × 2 m grid within 10 min and a 2.5 m × 1 m grid within 6 min. This outcome shows that the proposed robotic method can drastically reduce the time of manual inspection. Furthermore, a severity map was generated using the visual image results. This map highlights areas that require greater attention for repair in the test grid.

Vision-Based Concrete-Crack Detection on Railway Sleepers Using Dense U-Net Model

“…They utilized channel-wise distillation knowledge to make the model lightweight. Yang [46]. The authors incorporated attention modules to increase the accuracy of the model without any computational head.…”

Visual Detection of Road Cracks for Autonomous Vehicles Based on Deep Learning

“…Developed [ 23 ] a cost-effective unmanned aerial vehicle (UAV) equipped with sensors like LIDAR and optics for inspecting concrete pavement structure cracks [ 7 , 8 ]. Utilizing the SVM algorithm for data classification, this UAV overcomes the challenges of real-time crack detection [ 9 , 10 , 14 , 15 , 17 , 18 , 19 ] within budget constraints. While the method’s computational efficiency is a major advantage for low-cost UAVs, the paper does not address how this efficiency might impact the accuracy or reliability of the inspection results [ 29 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 , 39 , 40 ].…”

“…However, the paper does not address any potential limitations of this system, such as constraints in operational environments or challenges in detecting certain types of cracks. Study [ 19 ] developed Panthera, a robotic platform for road crack segmentation and garbage detection. However, their study does not address the limitations of using the mobile mapping system (MMS) for geotagging defects, nor does it explore the technique’s scalability across larger or varied pavement types [ 40 , 41 , 42 , 43 , 44 , 45 , 46 , 47 ].…”

“…The main challenge [ 11 , 12 , 13 , 14 , 15 ] in detecting road cracks arises from the diverse types of cracks caused by severe weather and long-term vehicle use. Traditional manual detection methods are inefficient and often miss cracks, while machine learning (ML) [ 16 , 17 , 18 , 19 , 20 ] models require manually designed features for crack detection. According to the Federal Highway Administration (FHWA), there are various crack types like fatigue, longitudinal, and transverse, making it difficult to design a universal feature extraction model for effective automatic detection using machine learning.…”