|
Instructors: Dennis Becker; Connor Ulf Christoph Gäde
Event type:
Project
Displayed in timetable as:
MProj - HRI
Hours per week:
6
Language of instruction:
English
Min. | Max. participants:
- | 15
Comments/contents:
The course will be conducted in presence on campus. Details will be disseminated prior to course begin on STiNE.
The project centers on the topic of Human-Robot Interaction. As the name HRI already conveys - the fundamental aspects come from research areas such as cognition, learning, interaction, and robotics. Research in HRI usually centers on these questions:
What does it take for a robot to make us feel comfortable with it? How can such a robot learn independently or by interaction with a person? How can a robot interact with a person?
Inspired by the idea of having a robot at home, at work, or in public areas which helps with your everyday tasks, we want to define and solve tasks like e.g. a coffee-serving or informational robot in a home or office environment. As exemplary possible scenarios, you should have a look at the current RoboCup@Home scenarios. Even such seemingly simple tasks/interactions require a broad set of interconnected and robust capabilities like vision, communication, motion, and planning, and a good idea on how humans interact with each other. Within this project, we will focus on the interaction between humans and robot(s). The project may also include programming bio-inspired methods, interaction research, and implementing a robotic framework.
Learning objectives:
The project Human-Robot Interaction offers the opportunity to learn about the complexities of daily action, basic control of humanoid robots as well as the development of simple service robot applications. During the course of the project, special emphasis will be put on teamwork, i.e. collaborative problem solving. In this way, the students can learn and experience the special difficulties and chances of project development, from research and conceptualization via implementation and validation to presentation and demonstration. The project will prepare them for research and the industry in booming sectors such as Knowledge Engineering, Human-Robot Interaction and Robotics with both skills in computer science (programming and robotics) and soft skills (teamwork and the English language).
Didactic concept:
This year, the full project will be condensed into a single semester. In the first quarter, the integrated seminar offers the opportunity to dive into the topic and explore interesting approaches that could be implemented later on. Also, in the first half, the project will focus on learning how to use the available hardware and software as well as acquiring general concepts and some practical experience on biologically inspired robotics, robot control architectures, and project management. In the second half of the semester, the project will focus on the implementation and execution of a HRI scenario that will be chosen by the students. To facilitate development, the whole robot control can be divided into several modules and later integrated via a robot middleware such as ROS. We plan to make use of two full days each week until end of March where the students will have the opportunity for self-study or team-work in a modern research environment. Ultimately, the results will be documented and presented.
Literature:
• C. Bartneck, T. Belpaeme, F. Eyssel, T. Kanda, M. Keijsers, and S. Šabanovic, Human-Robot Interaction. Cambridge University Press, 2020-02. doi: 10.1017/9781108676649.
• I. Goodfellow, Y. Bengio, and A. Courville, Deep Learning. MIT Press, 2016.
• G. Hoffman and X. Zhao, “A Primer for Conducting Experiments in Human–Robot Interaction,” ACM Transactions on Human-Robot Interaction, 10(1), 2021, doi: 10.1145/3412374.
Material:
The initial working platforms will be the NICO humanoid robot and simulation software. These platforms have good abstraction layers from hardware and an API that permit us to focus on behavioural aspects of human-robot interaction and service robot applications. NICO (Neuro-Inspired COmpanion) is a child-sized humanoid robot that is designed to have human-like sensing and motor capabilities. The NICO robot was built by the Knowledge Technology group to be a research platform for multimodal human-robot interaction and neurocognitive models. Additionally, code and documentation developed in previous semesters will be made available to get started, use and extend as required.
|
