Efficient ROS 2 Integration and Planning Logic for Vehicle Intent Detection

Institut

Lehrstuhl für Fahrzeugtechnik (TUM-ED)

Typ

Semesterarbeit / Masterarbeit /

Inhalt

 

Beschreibung

The aim of this project is to integrate a tail-light detection model into the autonomous driving stack and evaluate its performance. 

MOTIVATION

Modular autonomous driving stacks excel at refining their behavior when new perceptual features are introduced. Currently, the open-source autoware.universe implementation lacks the ability to adjust planning behavior based on the tail lights or turn indicators of preceding vehicles.
The goal of this thesis is to close this gap by developing an efficient, deployment-ready ROS 2 component for our physical EDGAR research vehicle. You will bridge the gap between 2D machine learning outputs and 3D spatial planning, ensuring the vehicle can dynamically adapt its velocity or lane-change logic based on the intent of surrounding traffic.

Voraussetzungen

YOUR ROLE

The work packages of this thesis contain:

• ROS 2 Node Implementation: Develop a clean, efficient ROS 2 node (compatible with modern Autoware architectures) that processes incoming camera streams.
• 2D-to-3D Spatial Association: Map the 2D tail light detections onto the 3D bounding boxes tracked by the vehicle’s main LiDAR/radar perception stack.
• Intent-Aware Planning Logic: Design and implement a simple, robust logic layer within the behavior planning module to alter vehicle reaction patterns (e.g., proactive slowing when a lead vehicle's brake lights activate).
• Vehicle Integration & Testing: Deploy your pipeline onto the EDGAR research vehicle and conduct real-world validation and performance evaluation.

WHAT YOU SHOULD BRING ALONG

• High motivation and strong interest in autonomous driving and real-time software systems.
• Solid programming skills in C++.
• Previous experience with ROS 2 .
• Optional but Helpful Basic familiarity with machine learning deployment (e.g., ONNX Runtime or TensorRT) is beneficial.
Joint Work Note: This project is highly collaborative and will be executed in close cooperation with Mining and Training Pipeline for Tail Light and Turn Signal Detection in Autonomous Driving, who is building the data mining and training pipeline providing the 2D models you will deploy.

If you are interested in joining this project, feel free to submit your application along with your CV and transcript of records. We look forward to receiving your application!
 

This work is also available as IDP and can be done in English or German! 

EMail: loic.stratil<script>document.write('@');</script>

<noscript>(at)</noscript>tum.de, niklas.krauss<script>document.write('@');</script><noscript>(at)</noscript>tum.de 

 

Verwendete Technologien

PyTorch, Python, Detection, Active Learning, TUM EDGAR, Autonomous Driving

Tags

FTM Studienarbeit, FTM Krauss, FTM AV, FTM AV Safe Operation, FTM AV Perception, FTM Informatik, FTM Detection

Möglicher Beginn

sofort

Kontakt

Niklas Krauß
Raum: 3507
Tel.: +49172 1736882
niklas.krausstum.de