A new review published in Artificial Intelligence and Autonomous Systems highlights how artificial intelligence can tackle the common problem of missing traffic data in intelligent transport systems. This study classifies and compares key data assignment methods and provides a clear roadmap for researchers and urban planners to improve traffic management and smart city operations.
As cities around the world deploy more sensors and intelligent systems to manage traffic, hidden issues are undermining their efforts: missing data. Sensor failures, communication dropouts, and harsh environmental conditions often lead to traffic information gaps, complicating everything from real-time traffic light control to long-term urban planning.
In a comprehensive new review published on AIAS, researchers at Shandong Technology and Business University will explore the latest AI-powered technologies designed to automatically fill these data gaps. This paper, entitled “A brief review of the lack of traffic data in intelligent transport systems,” provides a systematic classification of existing methods and compares performance under various missing data scenarios.
“Incomplete traffic data can affect signal timing, congestion forecasts, and even emergency response planning,” said Kaiyuan Wang, lead author of the study. “Our goal was to provide a clear framework that helps you choose the right method for the right situation.”
In this review, we divide the data assignment technique into two broad categories. Structure-based methods that rely on essential low-rank structures and spatio-temporal patterns of traffic data. A learning-based method for learning complex data relationships using deep learning models such as GAN, GNN, and attention mechanisms.
“Structure-based methods are often more interpretable and work well with moderate deficiency,” explains corresponding author Dr. Xiaobo Chen. “However, if high rates or complex patterns are missing, learning-based methods, especially those using graph neural networks or generative models, will be even more powerful.”
The study also brings together published datasets such as PEM, Metr-LA, and TaxIBJ, as well as standard assessment metrics such as Mae, Mape, and RMSE to provide valuable resources for researchers considering benchmarking their models.
Perhaps most actually, the team tested multiple methods under unified conditions, developing decision workflows, allowing users to choose the best approach based on missing data types, rates, and available computational resources.
Despite these advances, challenges remain. “Real world traffic data is messy. It's not just randomly missing, it could be affected by traffic signals, weather, or time of day,” Wang says. “We also need a way to quantify prediction uncertainty quickly enough for real-time use.”
Looking ahead, the author highlights several promising directions, including the fusion of multi-source data, lightweight AI models for edge computing, and uncertainty-aware assignment techniques.
“We understand not only that AI fills up missing data, but why it is missing, and how to rebuild it to the fullest in ways that support safer and smarter cities,” adds Dr. Chen.
This review provides both academic resources and practical guides for everyone who works in smart transport, urban analytics, or AI-enabled infrastructure management.
This paper, “A brief review of the lack of traffic data in intelligent transport systems,” is now available on AIAS.
reference:
Wang K, Chen X, Xu N. A brief review of the lack of traffic data in intelligent transport systems. artif. Intel. Orton. syst. 2025 (2): 0006, https://doi.org/10.55092/iaas20250006
/Public release. This material of the Organization of Origin/Author is a point-in-time nature and may be edited for clarity, style and length. Mirage.news does not take any institutional position or aspect, and all views, positions and conclusions expressed here are the views of the authors alone.
