Autoware Mini Course
The goal of the course is to give an overview how a self-driving car works using Autoware Mini as an example. Autoware Mini is a minimalistic Python-based autonomy software developed in-house by UT Autonomous Driving Lab. It has been field-tested with our Lexus RX450h vehicle on the streets of Tartu. Passing this course should give you a good enough background to start contributing to Autoware Mini or to start using it in your research.
The course consists of two parts:
- Lectures and practicals introducing Autoware Mini.
- A project to implement a change in Autoware Mini from an idea to a pull request.
Lectures and practicals take place for the first 8 weeks every Tuesday 12:15-15:45 in computer class 2006 in the Delta building. The first half focuses on introducing an Autoware Mini module and the second half involves implementing this module in Python.
In the later project you choose an issue from the Autoware Mini Gitlab, implement a fix and get it through the code review. It is also meant to introduce you to the industry's best practices in code maintenance and software quality.
The full course is worth 6 ECTS. It is possible to also do only the first part (Autoware Mini lectures and practicals) and earn 3 ECTS.
Course schedule
Schedule for lectures and practicals:
| 13.02.2024 | Introduction to ROS | (slides) | (practice) |
| 20.02.2024 | Localizer | (slides) | (practice) |
| 27.02.2024 | Controller | (slides) | (practice) |
| 05.03.2024 | Global planner | (slides) | (practice) |
| 12.03.2024 | Obstacle detection | (slides) | (practice) |
| 19.03.2024 | Local planner | (slides) | (practice) |
| 26.03.2024 | Traffic light detection | (slides) | (practice) |
| 02.04.2024 | CARLA simulator | (slides) | (practice) |
Schedule for projects:
| 09.04.2024 | Choose a project (Gitlab issue) |
| 16.04.2024 | Submit an initial design (Gitlab issue comment) |
| 23.04.2024 | Get a review for the initial design (Gitlab issue comment) |
| 30.04.2024 | Implementation |
| 07.05.2024 | Implementation |
| 14.05.2024 | Code submitted for review (Gitlab merge request) |
| 21.05.2024 | Get code review feedback (Gitlab merge request comments) |
| 28.05.2024 | Final code submitted (updated Gitlab merge request, all comments closed) |
Prerequisites
You should be fairly comfortable in:
It is helpful if you know:
- ROS (Robot Operating System)
- General robotics, i.e. what are transforms
- General machine learning, e.g. nearest neighbors search
Deliverables
To get 3 ECTS for lectures and practicals you need to present:
- Working solutions (code) for all 8 practicals.
To get additional 3 ECTS for the project you need to present:
- Final code as a merge request with all outstanding code review comments closed.
Contacts
Course instructors:
- Tambet Matiisen, tambet.matiisen@ut.ee
- Edgar Sepp, edgar.sepp@ut.ee
They can be found in Autonomous Driving Lab (room 3095 in Delta, behind Sandbox).