Crowdsensing for Road Pavement Condition Monitoring: Trends, Limitations, and Opportunities

Abstract

Road Pavement Condition Monitoring (RPCM) is indispensable for proactive maintenance, especially amidst increasing traffic and unpredictable weather patterns. The demand for cost-efficient solutions leveraging emerging technologies such as the Internet of Things (IoT), Machine Learning (ML), and cloud computing is increasing. This work examines the evolution of RPCM solutions, examines the challenges, and proposes future improvements. An extensive literature review is presented which exposes the challenges with existing RPCM solutions. The assessment criteria are the sensory platform, algorithms employed, detected road deformities, and performance. The approaches employed in RPCM are examined including their advantages and limitations. A holistic assessment of RPCM methodologies is presented which includes threshold, dynamic time warping, computer vision, and ML approaches. It is determined that smartphone-based monitoring solutions incorporating data acquisition and ML are superior to other methods. Future research directions are presented considering the limitations of existing solutions and the goal of cost-effective and efficient RPCM solutions.

Keywords

• Cloud computing
• computer vision
• Internet of Things (IoT)
• machine learning
• road monitoring
• sensors