Overview
Autonomy, in the context of autonomous vehicles, is the ability of a vehicle to operate without the intervention of a human operator. Such vehicles constitute a very broad spectrum of shapes and sizes including, for instance, vacuum cleaning robots, automated guided vehicles (AGVs) operating on factory floors and warehouses, and robot taxis etc.
The objective of the course is to give students knowledge in the area of autonomous vehicles, by participating in an autonomous-vehicles project. This can either be an applied or a research project in the field of autonomous vehicles.
The course is a two-part course i.e. 3+3 ECTS, in two semesters.
Information about project types
The student is (in small teams, or independently) expected to work on the applied/research project agreed-upon with supervisor.
The projects, this semester, are broadly of two types:
1. The students have the opportunity of working on our own autonomy stack the Autoware Mini (https://adl.cs.ut.ee/lab/software). Currently at ADL, we are actively enhancing the capabilities of the software, and by taking up this type of project, you can not only familiarise yourself with the software but can also contribute to its development. (See the "Autoware Mini Course" tab for more details.)
2. The students can take up a research or applied project using the DonkeyCar 1:10 scale platfrom. For the first 5 weeks, the students will follow prepared materials and tasks on how to use the car and train a self-driving model that is robust to environmental changes. A shortened version of those materials and demo videos of the expected results can be seen in this notebook.
After the teaching materials have been completed, the students will develop projects making the vehicle perform tasks in a small-scale autonomous-vehicle toy city (e.g. traffic sign or traffic light detection, obstacle avoidance) or elsewhere in the Delta building (e.g. making the vehicle take the elevator, drive a long route from point A and B). (See "1:10 scale car track" tab for more details.)
The DonkeyCar platform
See some videos from the toy town here.
Deliverables
The assessment is project-dependent. More details will be provided by the project supervisors in the beginning of the semester.
Contact
Naveed Muhammad (naveed.muhammad@ut.ee)
Tambet Matiisen (tambet.matiisen@ut.ee)
Ardi Tampuu (ardi.tampuu@ut.ee)