Call for Papers

  • Conference: The IEEE International Conference on Intelligent Transportation Systems (ITSC) 2026
  • Date & Place: September 15–18, 2026, Naples, Italy
  • Submission site: Papercept Portal -> Submit a contribution to ITSC 2026 -> Invited Session Paper
  • Submission deadline: March 1, 2026
  • Session code: 62d7u

Note from the official ITSC 2026 website:

We encourage authors to consider submitting their manuscripts to an Invited Session, when possible, rather than to a Regular Session, as this can help ensure their work reaches the most relevant and engaged audience.

We look forward to your contributions and to an engaging session at ITSC 2026!

Motivation and General Scope

Connected and Automated Vehicles (CAVs) are shaping the future of Intelligent Transportation Systems (ITS). Reliable operation requires strong safety assurance, scalability, and real-time performance under dense traffic interactions and in the presence of heterogeneous road users. Learning-based methods are increasingly used in CAVs for perception, prediction, and decision-making, but they remain difficult to validate for safety-critical operation, particularly under distribution shift and rare events. Model-based optimal control offers a structured approach to constraint enforcement and safety analysis, yet it faces practical limitations related to modeling accuracy, nonlinearity, nonconvexity, scalability, and adaptation to complex traffic interactions. Bridging learning-based methods and optimal control is therefore essential for reliable CAV operation in ITS, including mixed autonomy settings, connected cooperation, and large-scale deployment.

Topics of Interest

We invite contributions on theory and applications in the following categories:

  • Safety-Enhanced Learning: control-theoretic methods to assure learning-based methods, including but not limited to
    • Constraint- and risk-aware learning: constrained Markov decision processes, chance constraints, safe/offline learning under data limitations.
    • Runtime safety assurance: safety filters and shields.
    • Verification and validation: falsification, scenario generation, coverage-driven testing.
    • Safety-aware perception/prediction/localization for multi-agent learning.
  • Learning-Enhanced Control: learning components that improve control-theoretic methods, including but not limited to
    • Learned models for prediction/control: system identification, residual dynamics, disturbance/uncertainty models, learned interaction predictors for traffic agents.
    • Learning-augmented predictive control: learned dynamics refinements, learned uncertainty sets/tubes, data-driven objective and constraint design, bilevel tuning, learning warm-starts and surrogate solvers for real-time deployment.
    • Multi-agent optimal control with learning components.

Organizers

Portrait of Jianye Xu

RWTH Aachen University

Portrait of Bassam Alrifaee

University of the Bundeswehr Munich

Portrait of Andreas Malikopoulos

Cornell University

Portrait of Amr Alanwar

Technical University of Munich

Portrait of Chen Lv

Nanyang Technological University

Portrait of Ehsan Hashemi

University of Alberta

Portrait of Javier Alonso-Mora

Delft University of Technology

Portrait of Johannes Betz

Technical University of Munich

Portrait of Rahul Mangharam

University of Pennsylvania

Contact

Jianye Xu: xu@embedded.rwth-aachen.de