Introduction to Autonomous Robots

(0 reviews)


Nikolaus Correll, University of Colorado at Boulder

Pub Date: 2016

ISBN 13: 978-1-4937730-7-7

Publisher: Independent

Read This Book

Conditions of Use



  All reviews are licensed under a CC BY-ND license.

Learn more about reviews.

There are no reviews for this book

Table of Contents

1 Introduction
1.1 Intelligence and embodiment
1.2 A roboticists' problem
1.3 Ratslife
1.4 Challenges of Mobile Autonomous Robots
1.5 Challenges of Autonomous Manipulation
2 Locomotion and Manipulation
2.1 Locomotion and Manipulation Examples
2.2 Static and Dynamic Stability
2.3 Degrees-of-Freedom
3 Forward and Inverse Kinematics
3.1 Coordinate Systems and Frames of Reference
3.1.1 Matrix notation
3.1.2 Mapping from one frame to another
3.1.3 Transformation arithmetic
3.1.4 Other representations for orientation
3.2 Forward kinematics of selected Mechanisms
3.2.1 Forward kinematics of a simple arm
3.2.2 Forward Kinematics of a Di erential Wheels Robot
3.2.3 Forward kinematics of Car-like steering
3.3 Forward Kinematics using Denavit-Hartenberg scheme
3.4 Inverse Kinematics of Selected Mechanisms
3.4.1 Solvability
3.4.2 Inverse Kinematics of a Simple Manipulator Arm
3.4.3 Inverse Kinematics of Mobile Robots
3.5 Inverse Kinematics using Feedback-Control
3.5.1 Feedback control for mobile robots
3.5.2 Inverse Jacobian Technique
4 Path Planning
4.1 Map representations
4.2 Path-Planning Algorithms
4.2.1 Robot embodiment
4.2.2 Dijkstra's algorithm
4.2.3 A*
4.3 Sampling-based Path Planning
4.3.1 Basic Algorithm
4.3.2 Connecting Points to the Tree
4.3.3 Collision Checking
4.4 Path Smoothing
4.5 Planning at di erent length-scales
4.6 Other path-planning applications
4.7 Summary and Outlook
5 Sensors
5.1 Robotic Sensors
5.2 Proprioception of robot kinematics and internal forces
5.3 Sensors using light
5.3.1 Reflection
5.3.2 Phase shift
5.3.3 Time-of-flight
5.4 Sensors using sound
5.4.1 Ultra-sound distance sensors
5.4.2 Texture recognition
5.5 Inertia-based sensors
5.5.1 Accelerometer
5.5.2 Gyroscopes
5.6 Beacon-based sensors
5.7 Terminology
6 Vision
6.1 Images as two-dimensional signals
6.2 From signals to information
6.3 Basic image operations
6.3.1 Convolution-based lters
6.3.2 Threshold-based operations
6.3.3 Morphological Operations
7 Feature extraction
7.1 Feature detection as an information-reduction problem
7.2 Features
7.3 Line recognition
7.3.1 Line tting using least squares
7.3.2 Split-and-merge algorithm
7.3.3 RANSAC: Random Sample and Consensus
7.3.4 The Hough Transform
7.4 Scale-Invariant Feature Transforms
7.4.1 Overview
7.4.2 Object Recognition using scale-invariant features
8 Uncertainty and Error Propagation
8.1 Uncertainty in Robotics as Random Variable
8.2 Error Propagation
8.2.1 Example: Line Fitting
8.2.2 Example: Odometry
8.3 Take-home lessons
9 Localization
9.1 Motivating Example
9.2 Markov Localization
9.2.1 Perception Update
9.2.2 Action Update
9.2.3 Summary and Examples
9.3 Particle Filter
9.4 The Kalman Filter
9.4.1 Probabilistic Map based localization
9.4.2 Optimal Sensor Fusion
9.4.3 Integrating prediction and update: The Kalman Filter
9.5 Extended Kalman Filter
9.5.1 Odometry using the Kalman Filter
10 Grasping
10.1 How to nd good grasps?
11 Simultaneous Localization and Mapping
11.1 Introduction
11.1.1 Special Case I: Single Feature
11.1.2 Special Case II: Two Features
11.2 The Covariance Matrix
11.4 Graph-based SLAM
11.4.1 SLAM as Maximum-Likelihood Estimation Problem
11.4.2 Numerical Techniques for Graph-based SLAM
12.1 Converting range data into point cloud data
12.2 The Iterative Closest Point (ICP) algorithm
12.2.1 Matching Points
12.2.2 Weighting of Pairs
12.2.3 Rejecting of Pairs
12.2.4 Error Metric and Minimization Algorithm
12.3 RGB-D Mapping

About the Book

This book introduces concepts in mobile, autonomous robotics to 3rd-4th year students in Computer Science or a related discipline. The book covers principles of robot motion, forward and inverse kinematics of robotic arms and simple wheeled platforms, perception, error propagation, localization and simultaneous localization and mapping. The cover picture shows a wind-up toy that is smart enough to not fall off a table just using intelligent mechanism design and illustrate the importance of the mechanism in designing intelligent, autonomous systems. This book is open source, open to contributions, and released under a creative common license.

About the Contributors


Nikolaus Correll is a roboticist and an Assistant Professor at the University of Colorado at Boulder in the Department of Computer Science with courtesy appointments in the departments of Aerospace, Electrical and Materials Engineering.