Automated and Connected Driving Challenges

عن

Automated and connected driving is a major topic in automotive research and industry at the moment. The MOOC "Automated and Connected Driving Challenges (ACDC)" introduces participants to some of the latest research challenges and provides the possibility to develop and test automated and connected driving functions step by step.

This course first provides a comprehensive introduction to the Robot Operating System (ROS), which is a popular software framework for automated vehicle prototypes. On this basis, participants then learn how to develop and integrate modules for sensor data processing, object fusion & tracking, vehicle guidance, and connected driving. In particular, this MOOC allows participants to

• develop functions for automated and connected vehicles using Python and C++;
• integrate their developed functions into the Robot Operating System (ROS);
• train neural networks for environment perception tasks using TensorFlow;
• learn how to use tools like: Linux, Terminal, Docker, ROS, RVIZ, Juypter Notebooks, Git.

At the end of the course, you may optionally choose from a provided list of open research challenges and start working on your own contribution to automated and connected driving.

What you will learn

After completing the course, you will be able to

• contribute to current research challenges in automated and connected driving;
• program functions for automated and connected driving using Python & C++;
• integrate your developed functions into the Robot Operating System;
• train neural networks, e.g. with TensorFlow;
• evaluate your developed functions.

Skills you learn

• Docker (Software)
• Robot Operating Systems
• Artificial Neural Networks
• Python (Programming Language)
• Data Processing
• Research
• Linux Console
• TensorFlow
• C++ (Programming Language)

Syllabus

Week 1-3: Introduction & Tools

• Introduction to current challenges in automated and connected driving
• Introduction to the course tools and setup
• Introduction to the Robot Operating System (ROS1 & ROS2 Outlook)

Week 4-7: Sensor Data Processing

• Introduction to Sensor Data Processing
• Semantic Camera Image Segmentation
• Semantic Point Cloud Segmentation
• Object Detection in Point Clouds
• Occupancy Grid Mapping using Point Clouds
• Camera-based Semantic Grid Mapping
• Vehicle Localization

Week 8-9: Object Fusion and Tracking

• Introduction to Object Fusion and Tracking
• Object Prediction
• Object Association
• Object Fusion

Week 10-11: Vehicle Guidance

• Introduction to Vehicle Guidance
• Navigation-Level
• Guidance-Level
• Stabilization-Level

Week 12-13: Connected Driving

• Introduction to Connected Driving
• Collective Cloud Functions
• V2I-Communication

Week 14-15: Final Exam Period

• We suggest you take between one and two weeks to recap the materials of the course and then to finish the exam. Of course, you may take the exam whenever you prefer, if you completed the course earlier than planned.

(Optional) Week 14+

• Self-paced work on an automated and connected driving challenge you may choose
• List of challenges, instructions, data, supporting materials are provided
• Challenges can be tackled alone or in groups
• Your results may be published on your personal GitHub page

Auto Summary

"Automated and Connected Driving Challenges" is an expert-level course offered by edX, focusing on the latest research in automated and connected driving. Participants will delve into key challenges and progressively develop and test driving functions. Ideal for science and engineering enthusiasts, the course offers professional and starter subscription options, providing an engaging and comprehensive learning experience.