E-learning course overview

This course will take place in remote teaching form. All interaction takes place through this webpage.

The course is open to all students. Please register by  follow the link to "e-learing". 

At the scheduled lecture times, we will have live life video-streamed lectures. Once you are registred, you will be able to see the link under "live sessions" in the e-learning portion of this web page. You can join to watch and listen. If you have a microphone, you can join with audio to ask questions. You video cameras will be disabled by default, as multiple video streams tax the clients of the video conference. (Browsers: Chrome and Firefox work best, Safari doesn't work so well, especially for watching the videos.)

Live sessions will start on Thursday, 23 of April, 2020 at 14:15. 

The lectures will also be recorded and will be available for asynchronous viewing, although participating in the live session and asking questions is highly encouraged. 

We will have live sessions for the exercises at the listed times. Th exercise sessions will not be recorded.  

All course material will be on this webpage, including links to all the video material, to the lecture slides, exercises, and readings.

After registering, you will be able to upload your solutions to the exercises and will also be able to see your results once your submission has been corrected. You will be also be able to use discussion forums to ask questions synchronously and interact with your peers. 

Autonomous robotics is an interdisciplinary research field in which embodied systems equipped with their own sensors and with actuators generate behavior that is not completely pre-programmed. Autonomous robotics thus entails perception, movement generation, as well as core elements of cognition such as making decisions, planning, and integrating multiple constraints.

This course touches on various approaches to this interdisciplinary problem. In the first half of the course, the main emphais will be on dynamical systems methods for generating movement in vehicles.  The main focus of the course is, however, on solutions to autonomous movement generation that are inspired by analogies with how nervous systems generate movement. In fact, the second half of the course will review core problems in human movement  science, including the degree of freedom problem, coordination, motor control, and the relex control of muscles. 

Lecturers

Details

Course type
Lectures
Credits
6 CP
Term
Summer Term 2020
E-Learning
e-learning course available

Dates

Lecture
Takes place every week on Thursday from 14:15 to 16:00 in an online live session in the e-learning course.
First appointment is on 23.04.2020
Last appointment is on 16.07.2020
Exercise
Takes place every week on Thursday from 16:15 to 17:00 in an online live session in the e-learning course.
First appointment is on 23.04.2020
Last appointment is on 16.07.2020

Exercises

The course is accompanied by exercises, which will be posted weekly. Participants will upload their solutions, which will be corrected and marked (for bonus points). The exercises and their solutions will be discussed by Rachid Ramadan in the weekly exercise live session.

Further reading

Readings will be posted on this web page. Also have a look at the web page of the Dynamic  Field Theory community that is interested in related problems and solutions. There you find more exercises, reading material, slides and lecture videos that have some overlap with the lecture.

Teaching Units

Introduction
Document Syllobus
Lecture slides Introduction to the course
Dynamical systems tutorial
Lecture slides Lecture Dynamical systems tutorial
Exercises Exercise for dynamical systems tutorial
Attractor dynamics approach for vehicle motion ("sub-symbolic")
Lecture slides Lecture on the sub-symbolic variant of the attractor dynamics approach to vehicle motion
Video Video Bifurcation on the robot

The implementation of the attractor dynamics shown in the lecture on a robot vehicle. The bifurcation from a central repellor to an attractor is demonstrated by showing movement to the target (cross marked on the ground behind the wall) when the opening between obstacles is varied in width. The work was performed by Pierre Mallet with Estela Bicho and the CNRS in Marseille in the 1990s. 

Video Robot vehicle moving to a target: entering enclosure

The target is the mark of the floor inside the enclosure. The robot "knows" where the target is in terms of the cartesian coordinates relative to its starting position and updates the heading angle toward the target by integrating its own motor commands. Obstacle avoidance is based on sensory information: a rough estimate of distance from the amount of infra-red light from an emittor that is reflected back. 

Video Robot vehicle moving to a target: entering enclosure via another path

The same situation as in the previous video, but now the path taken there is blocked by obstacles. The vehicle finds another path. 

Video Robot vehicle moving to a target: leaving an enclosure
Video Robot vehicle moving to a target: leaving an enclosure 2

Same as the previous video, but with the opposite initial orientation of the robot. 

Video Robot vehicle avoiding obstacles that are moving
Video Robot vehicle with obstacle avoidance and neural field for a target sound source

The target representation is based in dynamic neural fields (see Bicho, Mallet, Schöner download link) not discussed in these lectures, but the topic of my WS course. 

Video Robot vehicle with the second order dynamics doing phono-taxis

The video is a demo of the approach published by Bicho and Schöner (1997) download link

Document Background reading 1: the attractor dynamics approach for vehicle motion (sub-symbolic) [Use this for Exercise 3]
Document Background reading 2: Second order attractor dynamics approach to vehicle motion
Exercises Exercise 3 on the sub-symbolic attractor dynamics approach
Kinematics of manipulators and the degree of freedom problem
Lecture slides Lecture on the kinematics of manipulators and the degree of freedom problem
Exercises Exercice 5: Kinematics and UCM

The Institut für Neuroinformatik (INI) is a central research unit of the Ruhr-Universität Bochum. We aim to understand the fundamental principles through which organisms generate behavior and cognition while linked to their environments through sensory systems and while acting in those environments through effector systems. Inspired by our insights into such natural cognitive systems, we seek new solutions to problems of information processing in artificial cognitive systems. We draw from a variety of disciplines that include experimental approaches from psychology and neurophysiology as well as theoretical approaches from physics, mathematics, electrical engineering and applied computer science, in particular machine learning, artificial intelligence, and computer vision.

Universitätsstr. 150, Building NB, Room 3/32
D-44801 Bochum, Germany

Tel: (+49) 234 32-28967
Fax: (+49) 234 32-14210