Socially Intelligent Robotics

Doel vak

The overall objective of this course is that students will be able to apply basic design skills to create interaction designs for a social robot and develop key social skills for the social robot using AI techniques.

Students will gain an understanding about social robotics and related AI techniques to control a social robot (e.g., conversational AI, computer vision aimed at interacting with people, expressing gestures, etc.) and apply this knowledge by designing a social human-robot interaction with a robot in groups. (Applying Knowledge and Understanding)

Students will evaluate the social skills they developed for the robot (Making Judgments).

Students will present their design, prototype and evaluation results and communicate about their individual roles within their group (Communication).

Students will be challenged to take the initiative and direct their own learning by identifying relevant social skills for the robot given a selected interaction context, creating a design rationale that is grounded in existing (multi-disciplinary) literature, and subsequently implementing that in a robot (Learning Skills).

Inhoud vak

In this course we will take a user-centered approach to the design of social robots and look into AI techniques for developing a social robot that can interact with human users. We will look at the basic cognitive skills we expect a social robot to have, including visual perception (e.g. face recognition), speech recognition and dialogue, emotional expression through body language, and the architecture for integrating these various skills to execute them on the robot.

Aanvullende informatie onderwijsvormen

Lectures and practical group work. The lab sessions (approx 6-7 hours per week), selected guest lectures, midterm exam, and final presentations are mandatory.

Attendance regulations

• You can miss up to 2 lab sessions without consequences if you inform your group.
• These are your slip days you can use for any circumstances. Use them wisely. You do not need to use them.
• If you are more than 15 minutes late it counts as a slip day.
• If you do not have any slip days left, a 0.5 point penalty per missed day will be applied to your final grade.
• If you miss more than 4 lab sessions you fail the course.
• If you do not attend the first two lab sessions without informing your group, you will fail the course.

Toetsvorm

There will be a mid-term exam (20%) to assess the theoretical understanding of the concepts covered in the lectures and selected literature . This is an individual assessment.

Furthermore, you will be asked to complete a practical assignment with a group of students. Grades for each of the main deliverables will contribute to the final grade and will be weight as follows:

• Robot software (20%)
• Design document (50%)
• Final presentation (10%)

Every group member is expected to keep track of their weekly progress in the form of a logbook. In case we find clear differences in what and how much individual group members have contributed to the final result (deliverables), we may take this into account and differentiate grades for individual group members. We will also use input from the teaching assistants, who will discuss with your group each week to establish such differences.

A minimum grade of 5.5 for each component is required.

It will not be possible to redo the practical assignment (no resit).

Literatuur

A brief course manual will be made available. The main literature used will consist of existing literature (papers and other materials) on social robotics and related AI techniques.

Aanvullende informatie doelgroep

Overige informatie

Important: Students who cannot attend the practical sessions on campus cannot join this course (because these practical sessions are crucial for the learning goals).

GenAI use:

• GenAI will be part of the software used in the course.
• GenAI is allowed as a coding assistant, although we encourage students to create their own code.
• GenAI use is not allowed for creating any of the requested written documents.

Aanbevolen voorkennis

Students should have the programming skills (Python) and ability to learn to use a programming framework for social robots that will be made available in the course.