Automated Pavement Condition Evaluation Using 3D Technology and ML

Bio

Dr. James Tsai is currently a professor in the School of Civil and Environmental Engineering and also an adjunct professor in the School of Electrical and Computer Engineering at Georgia Tech. Dr. Tsai’s research focuses on the development of Spatial Information and Sensing Optimization (SISO) methodologies, concentrating on applications to roadway/infrastructure health and safety condition evaluation, prediction, and management with a special emphasis on pavements, sidewalks, and signs assets. The unique strengths of his research team include 1) sensing technologies, 2) data science and advanced computation method, including artificial intelligence and machine learning (AI/ML), computer vision, spatial data analysis and visualization. Dr. Tsai is a recognized worldwide leader in the automatic detection and classification of pavement distresses using emerging 3D laser technology, computer vision, and Machine Learning. Dr. Tsai has been competitively selected in the pioneering research projects with more than $3.5 million: “Remote Sensing and GIS-enabled Asset Management (RS-GAMS) Phases I and II,” (2010-2014), sponsored by the US Department of Transportation Research and Innovative Technology Administration (USDOT RITA) program. Dr. Tsai’s research over the past ten years has been largely responsible for 3D laser technology emerging as a mainstream technology in the US; a majority of transportation agencies have moved from manual visual inspection to automated pavement condition evaluation. His research project has been competitively selected for the 2017 AASHTO High-Value Research (HVR) Award (a national award) because of the successful implementation of 3D laser technologies and AI on a large-scale system. His ongoing research project (NCHRP 01-60) will develop and publish eight national technical standards, including the use of 3D printing for the verification of 3D technology, by the end of 2022. Since 2010, he has served as the Associate Editor of ASCE Journal of Computing in Civil Engineering.

Abstract

3D laser technology (transverse laser profilers) has become a mainstream technology in the US. One hundred percent of state DOTs have replaced traditional manual inspection methods for collecting 2D intensity and 3D range image data using 3D laser technology in support of automated pavement condition evaluation. This paper presents the unique characteristics of 3D pavement data, including the new US national technical standards on open-format of 3D pavement surface data and procedures for image data quality measurements. The applications in the automated detection and classification of pavement distresses and characteristics, including cracking, rutting, potholes, loss of aggregates, texture, etc. are also presented This paper presents our research study of “Remote Sensing and GIS-enabled Asset Management (RS-GAMS),” sponsored by the US DOT Research and Technology Innovation Administration (RITA) program, and our study of “Development of An Asphalt Pavement Raveling Detection Algorithm Using Emerging 3D Laser Technology and Macrotexture Analysis,” sponsored by the National Cooperative Highway Research Program (NCHRP) Innovation Deserving Exploratory Analysis (IDEA) program. The research supports the future of pavement maintenance and rehabilitation (M&R) planning, M&R treatment decision-making, and pavement construction quality control methods using computer vision and machine learning. The developing trend of automated pavement condition evaluation using Machine Learning will also be presented. The innovative crack fundamental element (CFE) model with a topological representation of cracks will be presented for characterizing and modeling large-scale, in-field infrastructure cracking properties to study real-world crack propagation behavior using the automatically extracted cracking with high granularity to advance the 3R decision-making (right treatment at right time at right location) for optimizing infrastructure management and for creating opportunities to build resilient and sustainable infrastructures using new materials and structural designs. Analogous to precision surgery in the medical industry, the concept of “precision maintenance” has been created and will be presented. Case studies will be presented to demonstrate the huge amount of maintenance savings for transportation agencies that can be created and realized with “precision maintenance.” The future “precision maintenance” can be supported with the automatically extracted pavement distresses with a high granularity that could not be achieved previously.