MIN-Fakultät
Fachbereich Informatik
TAMS
UHH > MIN > Informatik > TAMS > Lectures > WS 2022 > Projekte > Masterproject Intelligent Robotics (Part 1) Seitenende

64-850-P1 Masterproject Intelligent Robotics (Part 1)

Room: F-326
Date: Thursday, 12:15 - 15:45
Organizer: Michael Görner, Norman Hendrich

Videos


Old Demo Video (17 MB mp4).

Marimbabot Scenario

The project focuses on the development of a robotic system capable of playing musical notes and chords on a marimba. The system integrates various technologies, including optical music recognition, audio command processing, and precise motion control. It employs a custom vision model to read musical notes from a whiteboard and utilizes voice recognition to interpret human commands, and finally evaluates the robot’s performance with audio feedback. The project encompasses hardware components, including 3D-printed com- pliant mechanisms and servomotors, to mimic human-like mallet strikes. This multifaceted project demonstrates the fusion of robotics, music, computer vision, and audio processing, showcasing a robot capable of performing a complex musical instrument with precision. The entire project, including code and documentation, can be found on GitHub.

"Marimbabot" Master Project Report

Final project report (3.3 MB pdf).

Final project demo slides (3 MB pdf).

https://github.com/UHHRobotics22-23/MarimbaBot Github repository.

Integrated Seminar

The integrated seminar 64-850-S: Integrated Seminar Intelligent Robotics complements the project.

Appointments

The following schedule is tentative and will be updated according to the participants' previous knowledge and interests.

Tasks to be worked on in the project involve:

Content

Within this project you shall learn how to basically program a robot to do tasks like navigating, manipulating, interfacing with humans. Therefore ROS (robot operating system) as the standard open-source middle-ware will be used to program the robot.

Prerequisites:

Experiences in programming C/C++ and/or Python

Language:

English and German

Start:

Yearly starting in winter term.

Keywords:

Computergrafik, 3D, Robotik. Mensch-Maschine-Interaktion, Bildverarbeitung, Locomotion, Sensortechnik, Manipulation, Grasping, computer graphics, robotics, human-machine interface, image processing, sensor fusion, localization, navigation, gesture recognition

Learning goal

The goal of the Master project is basically two-fold. First, you get hands-on knowledge about various methods and approaches from different aspects of state-of-the-art robotics. Second, you learn about how to organize work and how to collaborate in a group towards your self-defined project goal.

Approach

You will be working as a group towards a specific robot application (demo) proposed and chosen by the students. The Master project runs over two semesters. In the winter semester (WS 2022), the project consists of two parts with individual STiNE numbers (64-580-P1 for the practical robot work and 64-580-S1 for the seminar part). The Master-project will be continued in the following summer term (4 SWS, no seminar part).

Successful participation includes literature review, self-acquiring of basic methods and concepts, development and tests of (robot) software, implementation of a prototype, presentation and documention of the results.

All reports including an oral presentation in the first semester will take place in the integrated seminar 64-850-S1.

Evaluation

To pass the course, students have to

Written Submissions

At the end of the project, the group has to document their efforts. This work is split in two parts.

First, all final source code and project structure that was developed during the project should be documented. The aim of this part is to enable a person with no affiliations to the project to run and develop on top of the final demonstration. Useful documentation includes a technical overview of each part (usually in the form of a README.md for each component) with sufficient instructions to start and develop the demonstration scenario. If there are known issues that can break demonstrations, these should be documented as well. Lastly, the actual source code should be sufficiently structured and commented to be accessible later on.

Second, all other aspects of the project should be summarized by the whole group in a document in the style of an IEEE conference publication of 6-10 pages in total. It should comprise everything you would expect to see in a scientific publication while abstracting away implementation details (which are documented in the first part already). Relevant sections might include