Kensbock, Robin (University of Luebeck), Bouzidi, Mohamed-Khalil (Free University of Berlin, Continental AG), Schildbach, Georg (University of Luebeck)

Vehicle As a Sensor - Dynamic Games for Occluded Vehicle Estimation on Highways

Scheduled for presentation during the Regular Session "S43b-Multi-Sensor Fusion and Perception for Robust Autonomous Driving" (FR-EA-T43), Friday, November 21, 2025, 13:50−14:10, Stradbroke

2025 IEEE 28th International Conference on Intelligent Transportation Systems (ITSC), November 18-21, 2025, Gold Coast, Australia

This information is tentative and subject to change. Compiled on October 18, 2025

Abstract

The performance of autonomous driving systems is fundamentally constrained by the information available about the environment. This can become a problem when facing inherent sensor limitations, such as occluded traffic participants. Inspired by the anticipatory behavior of human drivers, we introduce a model-based estimation method that leverages the interaction dynamics between vehicles to infer the position of an occluded leading vehicle based solely on the behavior of a visible follower vehicle. Drawing on the theory of dynamic games, we interpret observed deceleration patterns as the result of Nash equilibrium strategies, effectively turning the visible vehicle into a sensor. Our method consistently outperforms a rule-based baseline in safety-critical scenarios with short time-to-collision, and enables millisecond-level inference, making it suitable for real-time applications.