Monday 03 March 2025
A novel approach to real-time short-term traffic prediction has been developed by a team of researchers, utilizing distributed fiber-optic sensing (DFOS) and data assimilation techniques. The method, which combines cutting-edge technologies in the fields of optics and artificial intelligence, is designed to provide accurate predictions of traffic flow on expressways.
The problem of traffic congestion is a pressing issue worldwide, with millions of hours lost each year due to gridlock. To mitigate this problem, intelligent transportation systems (ITS) have been implemented, which rely on real-time data to optimize traffic flow and reduce congestion. However, traditional methods for collecting traffic data are often limited by the sparsity and quality of the information gathered.
DFOS technology offers a potential solution to these limitations. By embedding fiber-optic cables in roadways, it is possible to gather high-resolution data on traffic flow at a much higher frequency than traditional sensors. This data can then be used in conjunction with sophisticated algorithms to predict future traffic conditions.
The researchers’ approach combines DFOS with a stochastic cell-automata-based model for traffic simulation. The model uses mean velocities derived from the DFOS data as input, along with initial conditions estimated from the latest microscopic traffic state. This information is then used to simulate the evolution of traffic flow over time, allowing for accurate predictions of congestion length and travel time.
The method was tested using real-world data collected on two Japanese expressways, Ken-O Expressway and Odawara-Atsugi Road. Results showed that the proposed approach outperformed traditional methods, with mean absolute errors (MAEs) ranging from 10 to 15 kilometers per hour for predicted mean velocities, and significant reductions in error for congestion length and travel time predictions.
The potential applications of this technology are vast. By providing accurate and timely information on traffic conditions, it could help reduce congestion, decrease travel times, and improve air quality. Additionally, the use of DFOS technology could be extended to other areas, such as monitoring infrastructure health or detecting natural disasters.
While there is still much work to be done in refining this approach, the results presented by the researchers are promising. The integration of DFOS and data assimilation techniques has the potential to revolutionize the field of traffic prediction, and could play a key role in shaping the future of intelligent transportation systems.
Cite this article: “Real-Time Traffic Prediction Using Distributed Fiber-Optic Sensing and Data Assimilation Techniques”, The Science Archive, 2025.
Distributed Fiber-Optic Sensing, Traffic Prediction, Real-Time Data, Artificial Intelligence, Intelligent Transportation Systems, Traffic Congestion, Roadways, Highway Management, Data Assimilation, Stochastic Modeling.
