Direkt zum Inhalt springen

Master's Thesis: Adaptive Cooperative Perception for ETSI-Based Vehicle-to-Infrastructure Systems Using Covariance-Aware Movement Clustering

11.06.2026, Abschlussarbeiten, Bachelor- und Masterarbeiten

The Chair of Robotics, Artificial Intelligence, and Real-Time Systems offers a Master’s thesis focusing on adaptive cooperative perception for ETSI-based Vehicle-to-Infrastructure systems using covariance-aware movement clustering and track-to-track fusion. The thesis investigates how object information from CAM and CPM messages, including covariances, velocities, and object parameters, can improve multi-agent object association and global tracking in urban traffic scenarios.

Motivation & Relevance

Urban intersections remain one of the most challenging environments for autonomous vehicles due to occlusions, limited visibility, and highly dynamic traffic participants. Cooperative perception systems can significantly improve scene understanding by enabling vehicles and roadside infrastructure to share object-level observations beyond their individual line of sight.

An existing research pipeline already supports ETSI-based cooperative perception using CAM and CPM messages, multi-agent simulation in CARLA, and uncertainty-aware object tracking. However, current approaches often rely on purely spatial association methods that do not fully exploit uncertainty information encoded in object covariances or motion consistency across agents.

Extending the framework with covariance-aware clustering and movement-consistent association enables more robust object tracking in dense and occluded urban scenarios, reducing ID switches and improving global scene consistency.

This thesis shall build upon the following repositories:

  • ETSI ITS ROS2 Messages: https://github.com/ika-rwth-aachen/etsi_its_messages
  • Shared Local Dynamic Map (S-LDM): https://github.com/DriveX-devs/S-LDM/tree/main

Project Description

In this thesis, you will extend the existing cooperative perception framework by developing covariance-aware object association and movement-adaptive clustering methods for ETSI CAM/CPM-based communication in a CARLA simulation environment.

The system currently supports:

  • CARLA-based cooperative perception
  • Multi-agent simulation workflows
  • ETSI CAM/CPM message generation
  • Uncertainty-aware object detection and tracking

Your task is to improve multi-source object association by integrating covariance-aware clustering mechanisms that jointly consider spatial distance, velocity consistency, and uncertainty structure of detected objects. In addition, roadside infrastructure (RSUs) will be incorporated as additional perception nodes providing complementary viewpoints.

Your system will:

  • Process CAM and CPM messages containing object states and covariance matrices
  • Perform covariance-aware object association across vehicles and RSUs
  • Apply movement-consistent clustering for stable multi-agent tracking
  • Fuse distributed object tracks into a global consistent world model
  • Evaluate robustness improvements in occluded and dense urban scenarios

The result will be an infrastructure-augmented cooperative perception framework capable of more reliable object association and tracking in complex urban traffic scenes.

Your tasks

  • Extend the CARLA cooperative perception pipeline with covariance-aware object association methods
  • Develop movement-aware clustering for CAM/CPM object fusion
  • Implement uncertainty-aware similarity metrics using covariance matrices
  • Design and implement track-to-track fusion for multi-agent observations
  • Integrate RSU-based perception nodes into the cooperative system
  • Evaluate uncertainty-aware cooperative perception performance in CARLA

Your Profile

  • Master’s student in Computer Science, Robotics, Electrical Engineering or a related field
  • Proficiency in Python or C++
  • Strong interest in computer vision, autonomous driving, or probabilistic robotics
  • Experience with ROS2, CARLA, or multi-object tracking systems is beneficial

What you will gain

  • Expertise in cooperative perception and multi-agent tracking systems
  • Hands-on experience with ETSI CAM/CPM communication frameworks
  • Deep understanding of uncertainty-aware perception and covariance-based fusion
  • Practical experience with CARLA-based autonomous driving simulation
  • Insight into intelligent transportation systems and infrastructure-assisted perception

How to apply

Please send your CV and a transcript of your grades with your application.

Kontakt: erik-leo.hass@tum.de