Transportation Research Part B: Methodological
Call for Papers for forthcoming special issue: Methodological Advances for Contextual Traffic Management: Enhancing Decision-Making Through Side Information
Detailed scope of the special issue
The rapid proliferation of sensing technologies and urban data infrastructure has introduced new opportunities for using contextual (side) information in traffic management, such as weather conditions, incident reports, social events, and GPS trajectories. This auxiliary data can reveal valuable correlations with key uncertain outcomes, like travel demand, road capacity, or congestion level. For example, real-time weather forecasts can inform traffic signal plans to mitigate rain-induced congestion, or social media streams can signal unusual traffic congestion due to spontaneous public gatherings. Such information, when effectively utilized, has the potential to significantly improve traffic system responsiveness, robustness, and efficiency.
Recent advances in machine learning and operations research have led to the emergence of contextual optimization methods, which explicitly leverage side information to tailor decisions to the current context. These include frameworks such as improving the optimization model to incorporate better predictive information (prescriptive framework), integrating the operation cost function into the forecasting model (predictive framework), or directly optimizing the mapping from side information to decisions (various decision rules). In traffic systems, these paradigms offer the ability to go beyond generic rule-based control and instead prescribe decisions, such as dynamic routing or signal control that are finely tuned to contextual cues. Despite their potential, several challenges remain. These include model interpretability, computational tractability for real-time decision-making, integration with legacy infrastructure, and the reliability of side information sources.
This special issue will delve into pioneering theoretical frameworks that explore new mathematical formulations and algorithms for contextual decision-making in traffic systems. Specifically, we welcome work that formalizes how side information, such as weather data and public transit data, can be systematically integrated into traffic control policies. For instance, one promising direction is the development of contextual stochastic optimization models that use covariate-dependent uncertainty representations, enabling more accurate and responsive strategies for dynamic traffic signal control or demand-aware pricing. Another avenue is the design of closed-loop learning-optimization pipelines, where machine learning models are trained not just for predictive accuracy but for minimizing downstream decision costs (e.g., total delay or fuel consumption). Additionally, theoretical work on contextual robust optimization, which can provide performance guarantees under covariate shift or data noise, can play a crucial role in ensuring reliability of the decision. Contributions may also explore the trade-offs between interpretability and efficiency in contextual policies, the development of generalization bounds for decision models using limited contextual data, or the design of novel surrogate loss functions for end-to-end learning. In addition to theoretical advancements, the special issue seeks impactful applications that demonstrate the real-world utility of contextual optimization in traffic management.
Timeline
Submission deadline: March 1, 2026
Decision deadline: Dec 15, 2026
Special Issue Guest Editors:
Maged Dessouky, PhD (maged@usc.edu) Daniel J Epstein Department of Industrial & Systems Engineering, University of Southern California, United States
Ning Zhu, PhD (nzhu25@ustc.edu.cn) School of Management, University of Science and Technology of China, China
Jinlei Zhang, PhD (zhangjinlei@bjtu.edu.cn) School of Systems Science, Beijing Jiaotong University, China
Lixing Yang, PhD (lxyang@bjtu.edu.cn) School of Systems Science, Beijing Jiaotong University, China
Maged M. Dessouky
Tryon Chair in Industrial and Systems Engineering
Professor & Chair
Daniel J. Epstein Department of Industrial and Systems Engineering
University of Southern California
Los Angeles, CA 90089-0193
phone 213 740 4891
fax 213 740 1120
email: maged@usc.edu
sites.usc.edu/maged
-------------------------------------------