Time/Location: Thursday 4:30-7:10p
ST2, Room 12
Instructor: Jana Kosecka
Office hours: 2-3pm Wednesday
Contact: Office 417 S&T II, e-mail:
kosecka@gmu.edu, 3-1876
Course web page:http://www.cs.gmu.edu/~kosecka/cs685.html
The course will cover basic principles of design and practice of intelligent robotics systems. We will cover algorithms for the analysis of the data obtained by vision and ultrasound sensors and the design of reactive control strategies which comprise basic capabilities of the mobile robot. Notion of robot's configuration space and geometric and topological representations of the environment will be introduced and followed by different motion planning techniques for searching these continuos and discrete state space representations. Issues of integration will be examined in the context of different modeling paradigms for representing spatial properties, reasoning about environment and decision making with the emphasis on navigation, pose maintenance and exploration. Examples of these include hierarchical deterministic models of interacting finite state machines, hybrid systems models and more recently popularized probabilistic models. We will study and formulate interesting robotics tasks and show how they can be accomplished by individual robot or cooperative robot teams (such as flocking, foraging as well as robotic soccer).
The topics and techniques covered are relevant for students interested in robotics, computer vision, artificial intelligence as well as modeling and programming of complex distributed embedded systems which interact with dynamically changing environments. Programming aspects in the context of these types of systems will be also part of the course.
The course will comprise from lectures by the instructor and discussion and presentations of the selected research publications by students. The grade will be predominantly based on participation in class, discussions, presentation and projects. The projects will involve implementation of a systems in a mobile robot simulator and/or the actual mobile robot.
Prerequisites:
CS 580 Artificial Intelligence
optional prerequisites - Computer Vision,
Analysis of Algorithms
Students taking the class should be comfortable with linear algebra, calculus
and probability
Textbook:
R. Siegwart and I. Nourbakhsh: Introduction
to Autonomous Mobile Robots, MIT Press, 2004
Grading:
Homeworks 35 %
Exam 35% (there will be one
exam, last week of October or first week of November)
Project 30%
The homeworks should be handed in on time. Late submissions are accepted
(as opposed
to no submissions) but will incur late submission penalty.
Other recommended books:
Tentative List of Topics:
|
|
|
Rigid Body
Motion, Mobile Robot Kinematics
|
|
Sensing
(ultrasound, vision, GPS)
|
|
Modeling
of the environment
|
|
Map Building
|
|
Pose Estimation
|
|
Motion Planning
|
|
Reactive
Control
|
|
Representing
and reasoning about space
|
|
Propabilistic
reasoning about space and action
|
|
Control and
Learning in Robotics Systems
|
|
Cooperative
Robotics
|
|
Architectures
and Modeling Frameworks
|