SOW-MKI57
Developmental Robotics
Course infoSchedule
Course moduleSOW-MKI57
Credits (ECTS)6
Category-
Language of instructionEnglish
Offered byRadboud University; Faculty of Social Sciences; Artificial Intelligence;
Lecturer(s)
Coordinator		prof. dr. W.F.G. Haselager
Other course modules lecturer
Examiner		prof. dr. W.F.G. Haselager
Other course modules lecturer
Lecturer		prof. dr. S. Hunnius
Other course modules lecturer
Contactperson for the course		L.E.C. Jacques
Other course modules lecturer
Lecturer		dr. G.E.H. Ras
Other course modules lecturer
Academic year2017
Period
PER3-PER4  (05/02/2018 to 13/07/2018)
Starting block
PER3
Course mode
full-time
Remarks-
Registration using OSIRISYes
Course open to students from other facultiesYes
Pre-registrationNo
Waiting listNo
Placement procedure-
Aims

After the course, the student has obtained:

  1. Practical skills in designing and implementing robotics systems (programming the NAO humanoid robot).
  2. Skills how to design systems and interfaces that are more effective and intuitive based on the principles of perception and cognition.
  3. Knowledge of principles of cognitive science that can be used to design new, emerging technologies.
  4. Knowledge on how studies in cognitive science help in designing and implementing natural and more efficient robotics systems.
  5. Knowledge of the role of perception, action and planning in designing robotic systems.
  6. Knowledge about the state-of-the-art in cognitive developmental robotics topics.

Furthermore, the student will have developed skills and gained experience with respect to modern robotic hard- and software, have designed a computational model that mimics development of cognitive skills, implemented the model on a robotics setup and perform robot experimentation to detect the model’s strengths and potential flaws. Lastly, other skills have been strengthened, including team skills, presenting and discussing research concepts and ideas, critical reading and more.

Content

One of the greatest challenges in creating more intelligent artificial systems is to understand how human’s cognitive abilities emerge through learning and development. You’ll develop computational models that mimic the development of cognitive processes in children. A great deal about your model’s strengths and potential flaws can be learned through its instantiation in a system embedded in the real world. In our Robotics Lab, you’ll be able to use state-of-the art humanoid robots (i.e. the NAO robot) as testing platforms for your models. Robot experimentation will give you a better understanding of how human and artificial cognitive processes can emerge through interaction with the physical and social environment.

Course topics:

  1. Principles of cognitive developmental robotics
  2. Varieties of embodiment
  3. Intrinsic motivation models
  4. Visual object processing  
  5. Motor-skill acquisition
  6. Theory of mind for robots
  7. Designing sociable robots
  8. Language learning
  9. Reasoning with abstract knowledge
Levels
AI-MA

Test information
The course is predominately lecture-based with in-class participatory mini-projects, group homework assignments, a group term project, and a final exam. Grades will be assigned 10% for class participation (including various “mini-projects”), 12.5% for each of the four homeworks, 20% for the term project, and 20% for the final exam.

Students will be divided into small groups (3-4 persons) to work on the homework and projects collaboratively. In principle, the groups should stay the same over the duration of the course. The groups are expected to meet five milestones. Four homework assignments (in steps) will prepare the experimental setup and the final project will use the system and methodology prepared during the homework assignments. For the final projects, students will develop a computational model that replicates a finding from human empirical studies on cognition. There will also be occasional “mini-projects,” which students will need to do on their own and that will count towards the class participation portion of the grade.

All homework assignments and the final project include presentations and writing. For the final projects students will write a research paper. Each group should submit a single paper, and all group members should contribute equally to the paper. Each homework will also have short (15 min) presentations by each group followed by a discussion session, which will count towards the grade for that assignment. The final project will have a 20 minute presentation for each group, also followed by a discussion session.

Prerequisites
BSc Artificial Intelligence or Computer Science (other: such as Sociology/Psychology: contact the course coordinator).

Contact information
Dr. W.F.G. Haselager; E: w.haselager@donders.ru.nl; T: 024-3616066

Recommended materials
Book
Cangelosi, A., and Schlesinger, M., Developmental Robotics. From babies to Robots, MIT Press, 2015. Additionally, relevant pointers to the literature will be given throughout the course.
Syllabus
The course syllabus (to be distributed and updated during the course) will describe all papers that will be discussed in the class.

Instructional modes
Assignments

General
Course work involves reading relevant literature, presentations, several medium-sized reports, four group assignments, and a practical project running throughout part of the semester. Some of the assignments, including the main practical project, involve applications and experiments on a real (Nao) robot. In addition, other (types of) robots may be used throughout the course.

Discussiongroups

Lecture

General
Weekly meetings combining formal lectures, student presentations and demonstrations, possible guest lectures, and literature discussions.

Presentation

Tests
Exam
Test weight1
OpportunitiesBlock PER3, Block PER4