In our pursuit of excellence, we continually set benchmarks, ensuring that our research remains at the forefront of both academic and practical applications in road infrastructure.
Use of Non-Destructive Technologies for Civil Infrastructure Health Monitoring
Keywords: Ground Penetrating Radar (GPR), Machine Learning Algorithm, Pavement Structural Deterioration
Our group has established a Ground Penetrating Radar (GPR) System to look beneath the surface of roadways in a non-invasive manner at a highway speed (65 mph) and assess their condition beyond what the naked eye can see. With a developed machine learning algorithm, Kim lab group uses the scanned GPR data: 1) to detect sinkhole potential and 2) to predict the potential pavement structural deterioration due to the weak pavement foundation.
“Revolutionizing Subgrade Assessment with GPR”
In our first stage, we focused on developing a state-of-the-art subgrade density prediction model utilizing cutting-edge Ground Penetrating Radar (GPR) technology. Through meticulous laboratory tests and field validation tests. particularly with high plastic silt soil, we successfully fine-tuned this model. Our research unveiled a powerful correlation between GPR scan results and subgrade soil properties, demonstrating the potential for non-destructive and highly accurate evaluations in civil engineering projects. The subsequent field tests further validated our model’s precision and reliability, establishing it as a game-changer for efficient subgrade assessment.
“Deep Learning Redefining Pavement Integrity”
In the second phase of our groundbreaking research journey, we embarked on extensive Ground Penetrating Radar (GPR) scans of road surfaces. Our mission? To harness the incredible capabilities of deep learning to revolutionize subsurface crack detection with unparalleled efficiency. Leveraging cutting-edge synthetic data augmentation techniques, our study achieved an astonishing test Average Precision (AP) score of 0.769 for crack detection. This remarkable feat underscores the transformative potential of this technology. While our primary focus centered on subsurface crack identification, our research unveiled an exciting opportunity: the seamless integration of this advanced deep learning model with subgrade density estimation from our pioneering first stage. This innovative approach promises to redefine pavement condition assessments, empowering decision-makers with invaluable insights to optimize infrastructure maintenance and rehabilitation strategies. Ultimately, our research sets the stage for elevated infrastructure durability and safety.
“Validation: The Future of Infrastructure“
In the final chapter of our research program, we embarked on a mission to validate our cutting-edge deep learning model developed in Stage 2. This model, designed for subsurface crack detection using Ground Penetrating Radar (GPR) scans of road surfaces, has already demonstrated exceptional accuracy with an impressive test Average Precision (AP) score of 0.769. Our validation results solidify its potential to revolutionize pavement condition assessments by providing timely and precise information about subsurface cracks. Moreover, our research underscores the thrilling prospect of integrating this deep learning model with subgrade density estimation from Stage 1, creating an innovative approach that promises to optimize infrastructure maintenance and rehabilitation strategies. This multifaceted research effort sets the stage for the future of infrastructure excellence, enhancing durability, and safety in the dynamic realm of civil engineering.
Remote Sensing Applications for Road Condition Monitoring by Means of Multispectral Satellite Imagery
Keywords: Pavement Monitoring, Remote Sensing, Multispectral Imagery, Machine Learning, Predictive Maintenance
Our lab leads cutting-edge advancements in pavement monitoring, vital for the socio-economic fabric of nations. Traditional methods, like PCI and IRI, have resource-intensive constraints. We’re leveraging remote sensing technology. Remote sensing captures data through electromagnetic radiation across the spectrum. In pavement engineering, we employ multispectral imagery, offering wide bandwidth and long-range capabilities. However, precision remains a challenge due to mixed pixel issues. To address this, we integrate machine learning with remote sensing. Using satellite data, our algorithms automate pavement condition analysis, streamlining predictive maintenance. In preliminary research, we analyzed concrete and asphalt pavements with multispectral and hyperspectral imaging. Concrete pavements showed consistently higher reflectance. Satellite imagery detects maintenance interventions, and hyperspectral scans reveal aging asphalt’s spectral changes due to traffic. Join us in advancing pavement monitoring for safer, cost-effective roadways.
Remote Sensing Applications for Road Condition Monitoring by Means of Multispectral Satellite Imagery
Road networks play a crucial role in socio-economical status of nations. Road conditions such as pavement aging and pavement damages affect rideability, safety and user costs. Maintaining high quality for road pavements is a crucial task for highway agencies. To that end, conditions of roads should be regularly assessed to grasp useful information about roads through time. Conventionally, road condition assessment has been done by visiting and inspecting roads. Inspectors use different indexes such as pavement condition index (PCI), International roughness index (IRI), and structural index to determine condition of roads. However, due to hazards related to these labor and time intensive in-situ visits, this method has made many problems. To overcome this issue, remote sensing that had been utilized in various fields, showed potentials to be used in pavement engineering by the mean of monitoring techniques.
To monitor road quality, either as a supplement or as an alternative to the ground-based methods, the application of remote sensing has been investigated by several researchers in last two decades. Remote sensing methods use electromagnetics radiations in different regions of electromagnetic spectrum to obtain information about the target object. In pavement engineering, one of the most common means of remote sensing is multispectral imagery. Multispectral images comprise a limited set of electromagnetic channels or bands, possessing wide bandwidth that facilitates the capture of signals from very far distances. This characteristic permits the deployment of multiple multispectral sensors into the orbit for the acquisition of multispectral images. When compared to other land categories, monitoring of roads via multispectral satellite images demands a significantly high spatial resolution. With recent enhanced spatial resolution of the multispectral sensors, these platforms have attracted significant attention among pavement researchers interested in multispectral imagery. However, despite the use of very high-resolution images, models were unable to achieve very high accuracy and the mixed pixel issue remains a as the main challenge in using multispectral satellite imagery in road pavement monitoring. The integration of machine learning techniques with remote sensing has emerged as a promising approach for monitoring pavement conditions
By leveraging machine learning algorithms to analyze this data, it becomes possible to automate the detection, classification, and tracking of pavement conditions over time, as well as predict future deterioration. In the first phase of the project, the preliminary study was done on both concrete and asphalt pavement sections using both multispectral and hyperspectral images to track changes of spectral profiles of concrete and asphalt pavements over time. It was seen that overall, concrete pavements have higher reflectance values in all the spectrum than asphalt pavements. Figure 2 shows the reflectance values (Red band) of the studied asphalt pavement section obtained from a satellite platform from 2010-2022. As seen in this figure, maintenance work can be detected from satellite imagery over time. Also, the result of hyperspectral scanning showed that there will be changes in the spectral properties of asphalt pavements because of aging and degradation by traffic passes.
In the second phase, our group has developed a model to utilize multispectral satellite images and ML algorithms to detect the type of pavements, and the conditions of asphalt pavements in a very large highway facility. Using historical satellite data, the dataset was generated with over 5.8 million datapoints and the result of classification showed an accuracy 74% in detecting road type and conditions.
Investigation of Heavier-Than-Expected Vehicle Weights Observed in the Vicinity of the Savannah Port Area and Their Impact on Georgia’s Pavements and Bridges and Rate of Statewide Asset Degradation
Keywords: Heavy Vehicle Weight, Pavement Degradation Prediction, Bridge Degradation Prediction
The Port of Savannah is one of the busiest maritime hubs in the U.S., experiencing escalating container capacity, which has put significant strain on local infrastructure. The research aims to quantify the impact of heavy vehicle traffic on pavements and bridges. Further, it seeks to assess the frequency of over-limit vehicles throughout Georgia, beyond just the port traffic. Weigh-in-Motion (WIM) technology will be utilized to gather the requisite data.
Investigation of Heavier-Than-Expected Vehicle Weights at Non-Interstate WIM Site in Georgia
The Port of Savannah, among the busiest maritime hubs in the U.S., has witnessed an escalating container capacity, significantly straining local infrastructure. This stress manifests as potential damage to pavements and bridges. This research focuses on quantitatively measuring the impact of heavy vehicle traffic on these structures and gauging the extent of damage. Additionally, there’s growing concern over vehicles in Georgia exceeding weight limitations beyond the Port’s traffic. This study seeks to quantify the frequency and weight of such over-limit vehicles. To achieve this, Weigh-in-Motion (WIM) technology will be employed at select Georgia sites.
1. Provide insights into the expected rate of transportation infrastructure degradation in Georgia, enabling GDOT to comprehend both the immediate and anticipated implications of heavy vehicle weights.
2. Bolster the reliability of infrastructure supporting the transport of goods near port facilities, thereby enhancing statewide economic development. This will be achieved through on-site examinations of pavements and bridge structures and an in-depth analysis of the assets informed by dynamic WIM data.
1. Examine the distinct characteristics of various non-interstate WIM sites in Georgia, leveraging data from GDOT’s Traffic Analysis & Data Application (GDOT-TADA).
2. Gather data concerning gross vehicle weight and its frequency of occurrence.
Weigh-in-Motion (WIM) system is specialized technologies designed to capture axle and gross vehicle weights as vehicles pass over sensors. This system can provide valuable data sets that include gross vehicle weight, axle loads, vehicle speed, and type.
In this research, WIM data from non-interstate sites in Georgia will be analyzed to assess the impact of heavy vehicle traffic on local infrastructure. Using data from the GDOT-TADA, the research aims to classify vehicles, calculate frequency based on gross vehicle weight, and determine ESALs(Equivalent Single Axle Load). This will help quantifying the stress on pavements and bridges, informing infrastructure planning and policy enforcement.
In Georgia, there are 30 active Weigh-in-Motion (WIM) sites. Of these, seven are situated on non-interstate roads, as illustrated in Figure 2. For this study, publicly available monthly WIM statistical data from 2021 to 2023 were collected from the GDOT-TADA website. This dataset comprises both directional and weight-based traffic volumes. Additionally, raw data for these seven non-interstate WIM sites were directly obtained from GDOT.
Using data collected from GDOT-TADA, the traffic characteristics of each WIM station were analyzed. Figure 3 represents one of the seven selected non-interstate WIM stations, specifically 115-0052. The graph is color-coded to differentiate between the years 2021, 2022, and 2023. The x-axis denotes the months, while the y-axis represents the average daily traffic volume. Given that the existing weight limit for vehicles in Georgia is 80kips and the weight limit for vehicles to and from the Port is 100kips, separate graphs were created to represent traffic volumes exceeding 80kips, between 80-100kips, and above 100kips. Additionally, both bidirectional and unidirectional (eastbound and westbound) traffic volumes were graphed. The key findings from the analysis are as follows:
• Consistent heavy vehicle traffic trends from 2021 to 2023.
• Eastbound traffic was more than twice as high as westbound.
• A higher proportion of vehicles exceeded 100kips.
The same graphical format was applied to analyze the traffic characteristics of the other six non-interstate WIM stations.
Using GDOT’s WIM raw data, we analyzed heavy vehicle traffic, focusing on vehicle classes 4 to 13 as categorized in Figure 4. An example analysis for WIM station 003-0132, based on 2021 data, is displayed in Figure 5. The left set of graphs shows the distribution of vehicle weight, single axle weight, and tandem axle weight for all target vehicle classes. For single-unit vehicles (Classes 4-7), the middle set of graphs displays similar distributions, while the right set of graphs covers multi-unit vehicles (Classes 8-13).
The vehicle weight distribution graph is color-coded based on load limits: the traditional 80 kips, 10% above standard (88 kips), and the new 100 kips limit. Single axle weight is color-coded according to reference points set at 18 kips for pavement and 20 kips for bridges. Tandem axle weight uses 34 kips as the standard reference point for both pavement and bridge, with additional markers at 25% above standard (42.5 kips) and 50% above standard (51 kips) also indicated.
Utilizing the dataset spanning from 2021 to 2023, the analysis incorporates all seven non-interstate WIM stations. The findings aim to illuminate the traffic characteristics unique to each station. An upcoming focus will be to quantify the degradation impact on both pavement and bridge structures due to heavy vehicles that exceed standard weight limits. Furthermore, plans are in place to correlate the weight violation frequencies with maintenance records to establish if higher rates of infractions lead to quicker degradation of infrastructure. Ultimately, this research intends to provide foundational data that can be used to refine transportation infrastructure policies, potentially influencing future road maintenance scheduling and budget allocation strategies.
Researches on Mechanistic-Empirical Pavement Design Guide
Keywords: Sustainable Pavement Materials, Mechanistic-Empirical Pavement Design (MEPDG) Manual, Forensic Analysis of Damaged Pavement, State Highway Agencies
Kim lab group has pursued researches to directly advance the art of sustainable pavement materials, design and maintenance. The area of Kim lab research is, but not limited to, in research and testing of pavement materials such as asphalt concrete, Portland cement concrete, aggregates, and soils to create truly sustainable pavements to build roadways at optimal costs. Dr. Kim has been a PI or Co-PI on numerous awarded external grants in this area. Based on the pavement researches funded by GDOT, Kim lab group has compiled all the pavement research outcomes into the GDOT Mechanistic-Empirical Pavement Design (MEPDG) Manual. In this design manual, Kim lab group provided guidance and recommendations to the Georgia engineers for utilizing the new MEPDG to more accurately design pavement structure and evaluate pavement performance. Kim lab group also developed a set of guidelines for the forensic analysis of damaged pavement that outline best methods for pavement repair and replacement for roadways. This step-by-step analysis enables the state highway agencies to identify problems early, fix the pavement distresses, and avoid costly delays.
Climate Resilient Natural Infrastructure – Funded by NSF/USDA NIFA
Keywords: Sea Level Rise, Soil Organic Carbon, Salt Marsh Resilience
As sea level slowly rises, many marshes can enhance C sequestration through sediment trapping and soil organic carbon (SOC) accumulation. SOC, often measured generally by proxy of soil organic matter (SOM), is a key parameter reflecting marsh accretion. With increasing rates of sea-level rise, some marshes have been unable to respond adequately, and the marsh platform becomes unstable; this phenomenon is projected to lead to reduced primary productivity, increased erosion, and eventually, mineralization of the stored SOM. Surface and belowground SOM is a critical soil property to salt marsh resilience. Surface and belowground SOM data in salt marshes obtained through remotely and continuous data collection in conjunction with field observations and predictions from machine learning (ML) algorithms will provide insight into the challenges faced by salt marsh research groups.