The course gives students with a special interest in robotics, perception and learning the opportunity to read an especially designed project course in their own interest field. Both course content and assessment are designed individually for each project.

Course setup

To be able to start the course, the student should first contact a teacher connected to the interest field and have them agree to supervise the project. The course can only be offered if the department has enough resources and skills in the current field. When this is done they can contact the course coordinator.

Step 0: Specification

The first step in the project work of the course is to write a specification for the project.

For the supervisor: If you compare this with a master thesis project this is somewhere between what you would write before the work even started and a detailed specification. The specification is meant to make both student and supervisor think carefully about how to set up the project before the student dives into it and determine what is expected as output. The specification has to provide enough detail to ensure that course coordinator can determine that it meets the requirements of the course (fits the topic, scope, examination, etc). It is also a tool to ensure that supervisor and student(s) have the same expectations as far as possible.

The specification should contain

description of goals of the project,
intended learning outcomes (ILOs) that can be access for each student,
a description of how the examination should be carried out for this specific project,
a description of roughly how the work is planned to make the students reach the ILOs.

The specification must be approved by the course coordinator. It is thus expected to be completed very early on in the project or before.

As in all projects a plan is just a plan and changes are likely. Communicate with your supervisor along the way and for major changes, consult the course coordinator.

Step 1: Project work and documentation

The project work can look very different from project to project. However, it is strong encouraged to break it down in such a way that there is a half time presentation to allow individual follow up to take place.

Typical deliverables at the end are an appropriate mix of final presentation, report and software / demonstrations. This should be documented in the specification.

Intended learning outcomes (ILOs)

Each project must define their own intended learning outcomes which must include the following.

Having passed the course, the student shall be able to

choose a course of approach and define, follow and follow up a plan for carrying out the task in a given resource budget,
Present orally and in writing, a description and defense of a technical solution to a problem in robotics, perception and learning

in order to

be able to participate professionally in project work in activities in robotics, perception and learning.

Examination

The setup of the course is

PRO1 - Project work, 7.5 credits, grading scale: P, F
PRO2 - Project work, 7.5 credits, grading scale: P, F

The examiner may apply another examination format when re-examining individual students.

A requirement to get a passing grade on the course is to provide feedback at the end.

Examples of previous projects

Fall 2025
- Autonomous Suction Grasp Planning (1 student)
- Modular Mobile VLAs by Combining Image-space Navigation with VLAs for Manipulation (1 student)
Spring 2025
- Steoro-vision based obstacle avoidance for an autonomous hydrofoiling craft (3 students)
- ARI as a Personal Office Assistant (4 students)
- Autonomous Outdoor Navigation Platform (4 students)
Fall 2024
- Interactive Multimodal Spot Robot Control System (1 students)
- Camera Radar fusion for Long-range 3D Perception (1 students)
- Adaptive Query Timing and Requested Information in Time-aware MOMDPs for Autonomous Decision-making (1 students)
- Foilcart RL: Reinforcement learning control of single-strut hydrofoils (2 students)
Spring 2024
- Altitude Estimation for eFoils (1 student)
- Corridor MPC with Human-in-the-Loop for Planar Quadrotors (1 student)
- Joint 3D object detection, speed estimation and tracking on multi-LiDAR data (1 student)
Fall 2023
- Multi-modal amodal instance segmentation (2 students)
- An unsigned distance field map representation (1 student)
Spring 2023
- Simultaneous localization and mapping in the forest (1 student)