Real-World Evaluation of two Cooperative Intersection Management Approaches

📅 2024-03-25
📈 Citations: 0
Influential: 0
📄 PDF

career value

189K/year
🤖 AI Summary
Existing autonomous driving coordination methods rely heavily on high automation penetration rates and idealized simulation assumptions, failing to address real-world mixed-traffic intersections with both autonomous and human-driven vehicles. Method: This paper proposes two cooperative, signal-free intersection management strategies enabling vehicle-infrastructure cooperative decision-making and heterogeneous vehicle co-driving. The approaches integrate multi-scenario traffic prediction, graph-structured reinforcement learning, and a unified software stack jointly deployed in both real-world vehicles and high-fidelity simulation. Contribution/Results: For the first time, the methods are jointly validated in real-road testing and mixed-traffic simulation—breaking reliance on pure simulation and high automation penetration. Empirical results show substantial improvements in intersection throughput: reduced crossing time and fewer stops. Notably, significant efficiency gains persist even at low autonomous vehicle penetration rates, with only marginal increases in safety-critical metrics. The core contribution lies in robust cooperative control under limited automation and a cyber-physical co-verification paradigm.

Technology Category

Application Category

📝 Abstract
Cooperative maneuver planning promises to significantly improve traffic efficiency at unsignalized intersections by leveraging connected automated vehicles. Previous works on this topic have been mostly developed for completely automated traffic in a simple simulated environment. In contrast, our previously introduced planning approaches are specifically designed to handle real-world mixed traffic. The two methods are based on multi-scenario prediction and graph-based reinforcement learning, respectively. This is the first study to perform evaluations in a novel mixed traffic simulation framework as well as real-world drives with prototype connected automated vehicles in public traffic. The simulation features the same connected automated driving software stack as deployed on one of the automated vehicles. Our quantitative evaluations show that cooperative maneuver planning achieves a substantial reduction in crossing times and the number of stops. In a realistic environment with few automated vehicles, there are noticeable efficiency gains with only slightly increasing criticality metrics.
Problem

Research questions and friction points this paper is trying to address.

Autonomous Vehicles
Traffic Management
Mixed Traffic Environment
Innovation

Methods, ideas, or system contributions that make the work stand out.

Connected Autonomous Vehicles
Intersection Traffic Flow Optimization
Reinforcement Learning Planning
🔎 Similar Papers
No similar papers found.
M
Marvin Klimke
Institute of Measurement, Control and Microtechnology, Ulm University, D-89081 Ulm, Germany; Robert Bosch GmbH, Corporate Research, D-71272 Renningen, Germany
M
Max Bastian Mertens
Institute of Measurement, Control and Microtechnology, Ulm University, D-89081 Ulm, Germany
B
Benjamin Volz
Robert Bosch GmbH, Corporate Research, D-71272 Renningen, Germany
M
Michael Buchholz
Institute of Measurement, Control and Microtechnology, Ulm University, D-89081 Ulm, Germany