Course: EW452 Advanced Topics in Robotics

3 Credits – 2 Recitation Hours – 2 Laboratory Hours

Course Description:

A follow-on to EW450 Introduction to Robotic Systems that introduces parameterizations, numeric inverse kinematics, path and trajectory planning, visual servoing, multiview camera calibration and depth recovery, visual features, probabilistic classification, neural networks, and fundamental convolution neural networks. Students develop methods for motion prediction, motion planning, closed-loop tracking, and target identification. Methods are applied during hands-on lab exercises and a multi-week final project using articulated robotic manipulators and machine vision cameras.

Pre-requisites:

EW450 or Dept. Chair Approval

Course Coordinator:

Assoc. Prof. Kutzer

Textbook:

None

Course Objectives:

1. Compute, apply, and manipulate rigid body transformations;

2. Derive, visualize, and explain parameterizations and the exponential map of rotations in space;

3. Derive, visualize, and explain parameterizations and the exponential map of rigid body transformations in space;

4. Derive and explain the world and body-referenced Jacobian using the exponential map in a coupled and decoupled context;

5. Derive, visualize, and explain numeric inverse kinematics, paths, and trajectories in n-dimensional space;

6. Derive and explain the Jacobian for fixed camera and eye-in-hand visual servoing of articulated robots including assignment of gain terms;

7. Utilize and explain multiview camera calibration techniques;

8. Derive, utilize, and explain depth recovery using multiview camera systems;

9. Utilize and explain visual features for object characterization; and

10. Utilize, and explain advanced methods for object/pattern recognition (Bayes Theorem, artificial neural networks, and fundamental convolution neural networks).

Topics:

1. Rotation Parameterization and the Exponential Map

2. Rigid Body Transformation Parameterization and the Exponential Map

3. Jacobians Leveraging the Exponential Map

4. Numeric Inverse Kinematics

5. Path Planning

6. Trajectory Planning

7. Visual Servoing

8. Multiview Camera Calibration

9. Depth Recovery Using Multiview Cameras

10. Quantifying Visual Features

11. Multivariate Probability Density and Probability Mass Functions

12. Naïve Bayes Classifiers

13. Fundamental Decision Theory

14. Neural Networks

15. Backpropagation

16. Fundamental Convolution Neural Networks