Course content

Over the last several years, ingenious forms and mechanisms found in nature have increasingly inspired robotic researchers to generate new solutions. For example, leg control strategies in stick insects have inspired control algorithms that allow not only insect robots to walk, but also a bipedal walking robot to walk with high speed. Gecko foot-hair has inspired new material which allows robots to climb up vertical walls.

This course consists of two main parts: (i) presentation of existing approaches in biologically inspired robotics and (ii) hands-on training on the design and implementation of bio-inspired robots. The topics covered in the course’s lectures include: fundamentals of traditional robotics, bio-inspired robot design (e.g. actuators, sensors, material), bio-inspired algorithms for robot control (e.g. neural network, evolutionary algorithm), practical robot building, etc.

The course is highly interdisciplinary and is targeted not only to students from computer science, but also to students of the faculty of mathematics and science (e.g. biology, neuroscience) and the philosophical faculty (e.g. psychology).

The course’s weekly lecture consists of presentation of fundamental concepts, case study discussions of existing bio-inspired robots, and in-class exercises. In-class exercises will provide hands-on experience with various bio-inspired design techniques and algorithms. These exercises will also allow students in a group of two to build up the necessary hardware and skills to implement a robotic system (as a class project). Throughout the course, each group will implement bio-inspired techniques of their choice on the robot, which will participate in a competition (e.g. predator vs prey robots) at the end.

Objective

In this course, students will (i) gain knowledge and understanding of the range of bio-inspired techniques that are applicable to robotic systems, (ii) learn how to implement bio-inspired techniques on robots (iii) learn basic skills in robot implementation and work together as a team to implement a robot, and (iv) gain experience with reading and discussing materials in research papers.

Grading

The course grade will be based on the open-book written exam (50%) and class project (50%).

Seven out of ten exercises have to be completed in order to participate in the exam.

Project

Each group of two people will build a robot for prey-predator task. All parts will be provided and in-class exercises will incrementally build on the robot's hardware and software systems.

At the end of project, there will be a prey-predator competition. Each robot has to serve both functions, sometimes as a prey and other times as a predator.

Project grading is based on three components:

- innovative ideas/solutions

- successful implementation

- clear presentation of both (ideas and implementation) -- in the form of a 4-page written report and 15-min presentation.

Exam

9 Jun, 14:00 - 16:00. The exam is open-book, i.e. any materials from or outside class can be brought.

Location: AND-3-02

Lijin Aryananda, email = lijin at ifi.uzh.ch

Office Hours: 15:00 - 17:00 every thursday

Literature

The course literature will be based on a series of research papers which will be provided on the class website and selected chapters from “Bio-Inspired Artificial Intelligence: Theories, Methods, and Technologies”, by Dario Floreano and Claudio Mattiussi, ISBN: 0262062712.

Lectures

Exercises