96994 - Smart Vehicular-Systems

Learning outcomes

Students taking this class: - Will An extended knowledge of challenges and opportunites in the area of vehicular communications including V2V and V2I. The class material will include real world examples and deployments. - Will understand impact of mobility and propagation on Vehicular Systems performances - Will learn vehicular application scenarios and their evolution - Will learn how design and model for connected/autonomous vehicular systems.

Course contents

Autonomy, Mobility and AI are the biggest innovations posed to affect our life in the years to come. Th

This course will provide a survey of mobile and autonomous systems with the goal of understanding the building blocks for autonomous mobile systems and their principle of operation. The course will explore a wide range scenarios to outline how in the current stage of technology is in its infancy and there is not yet a general solution.

Pre-requisites for the class is an excellent understanding of algorithms, AI concepts, and a good understanding of systems. The class is designed to be heavy on projects therefore good programming skills and ability to work with Linux and RT os are a must; furthermore a good knowledge of AI models, tools and techniques is preferential (i.e. Yolo, Tensor RT, etc). Knowledge of robot operating system (ROS), Nvidia CUDA/Jetson programming, and introductory level understanding of control systems are a plus.

The goal is to teach autonomous mobile systems fundamental tradeoffs and techniques to equip students with a deep understanding of how the autonomous vehicular systems have developed, and how they will evolve in the near future.

A successful student in this class will gain skills on:

• Understand the challenges and opportunities offered by advances in autonomous vehicles and robotics.
• Gain in-depth knowledge of advanced computer networking and telecommunications issues applied to autonomous connected vehicles including new network architectures beyond the Internet.
• Analyze the requirements for a given autonomous system and select the most appropriate system architecture and technologies.
• Gain an introductory level familiarity with tools for autonomous systems including simulation and practical systems.

 

Readings/Bibliography

The class material will be mostly provided by the instructor and will be based on recent research papers and classic introductory lectures on some topics. In addition Students will be required to take on-line tutorials for specific topics.

 

As reading list students can consider the following books:

1. O'Rourke, J. Computational Geometry, Cambridge University Press.
   
2. Kevin M. Lynch and Frank C. Park, Modern Robotics: Mechanics, Planning, and Control, Cambridge University Press.
3. Andreas Lindholm, Niklas Wahlström, Fredrik Lindsten, Machine Learning A First Course for Engineers and Scientists, Cambridge university Press.
4. Meyn, Sean, Control Systems and Reinforcement Learning, Cambridge university Press.

Teaching methods

The class will be taught by classic lectures, tutorials, on-line seminars, hybrid lectures, lab-exercises, and experimental work.

Lectures will be recorded for future reference.

Assessment methods

Team Project with report and presentation.

Teaching tools

On-line Tutorials, Microsoft Teams, Coursera short seminars, Open source simulators, research papers, slides.

Office hours

See the website of Giovanni Pau

See the website of Roberto Girau