(A) design of a scalable robotic system in the cooperative intelligent space

다국어 초록 (Multilingual Abstract)

This dissertation presents a scalable robotic system in the cooperative Intelligent Space, which is a spatial system that can be extended by the cooperation among
people, robots, and distributed sensors. The scalability of the space is important since more and more networked devices may be integrated and configured in the existing
system. Thus, in our study, we focused on a scalable and cooperative robotic system that is configured through cooperation between networked robots and distributed
networked sensors.
We describe a resource sharing architecture (RSA) for networked sensors and robots which are designed to realize the system. The most important characteristic of
RSA is that it allows networked robots to utilize external shared resources by providing an automated connection between the networked robots and external resources.
Networked robots need to determine their positions and orientations in order to carry out various and diverse tasks. We describe a motion-based identification scheme for networked mobile robots which are integrated and configured as part of the robot’s system. We describe a robot identification system that determines the mapping relation between the identity of the robot and its position, by analyzing the similarity between the given paths and trajectories of the robots in the Intelligent
Space using multi-camera networks.
As mobile sensors, the identified robots are able to perform environment mapping or sensor configuration using their coordinate systems. In the study presented in this dissertation, we utilize mobile robots to perform automated camera calibration as part of the configuration and extension of the Intelligent Space. To achieve this,
the odometry errors of mobile robots should be compensated, since the movement of the robots must be determined very accurately. After odometry error compensation has been performed, the position of the robot can be utilized as an absolute position in the world coordinate system.
Finally, we design and evaluate a scalable robotic system. For this purpose, we measure the localization error of the identified robots, their odometry errors, and the distance errors of 3D points reconstructed by calibrated cameras. In addition, we evaluate the RSA in terms of the configuration and extension of networked sensors and robots. Thus, the scalable robotic system facilitates automated configuration of networked devices in an extended space. In conclusion, we show that the scalability of the system allows the cooperative Intelligent Space to be extended and configured.

목차 (Table of Contents)

• Abstract i
• List Of Figures vii
• List Of Tables xi
• Chapter 1 Introduction 1
• 1.1 Motivation 1
• 1.2 Background 4
• 1.3 Contributions and organization of the dissertation 8
• Chapter 2 Scalable Robotic System in the Cooperative Intelligent Space 10
• 2.1 Overview of the Intelligent Space 10
• 2.1.1 The concept of the Intelligent Space 10
• 2.1.2 Properties of the Intelligent Space 10
• 2.1.3 Distributed Intelligent Networked Device (DIND) 11
• 2.2 Integration and Configuration of the Intelligent Space 12
• 2.2.1 Configuration of DINDs 12
• 2.2.2 Robot control based on the Intelligent Space 16
• 2.2.3 Design of framework for DINDs and robots 18
• 2.3 Problem statements 20
• 2.3.1 Architecture for DINDs and robots 20
• 2.3.2 Automation of preprocessors for robot localization 22
• 2.3.3 Movement of mobile robots for a scalable robotic system 23
• 2.4 Objective 26
• 2.4.1 Considerations 26
• 2.4.2 Requirements 27
• 2.4.3 System configuration 28
• 2.4.4 Architecture design 30
• Chapter 3 Improvement of Resource Sharing Architecture 34
• 3.1 Motivation and problem descriptions 34
• 3.2 Related work 36
• 3.3 Overview of resource sharing architecture 37
• 3.3.1 Structure of resource sharing architecture 37
• 3.3.2 Protocol interface using Jini 39
• 3.4 Design of an automated connection mechanism 41
• 3.4.1 Interface description for an automated connection 41
• 3.4.2 Functions for data communication 42
• 3.4.3 Functions for service component control 43
• 3.5 Architecture based on an automated connection mechanism 45
• 3.5.1 Configuration of architecture 46
• 3.5.2 Resource sharing mechanism 49
• 3.5.3 Error-handling mechanism 50
• 3.5.4 Load balancing mechanism 52
• 3.6 Simulation results 55
• 3.6.1 Performance analysis of component search 55
• 3.6.2 Performance analysis of load balancing 58
• 3.7 Discussion 63
• Chapter 4 Motion-Based Identification of Mobile Robots 64
• 4.1 Motivation and problem descriptions 64
• 4.2 Approach for motion-based robot identification 66
• 4.2.1 Definition and concept 66
• 4.2.2 Scheme for robot identification 68
• 4.2.3 System configuration 71
• 4.3 Feature extraction 72
• 4.3.1 Region of interest (ROI) segmentation 73
• 4.3.2 Color histogram feature 76
• 4.3.3 Criterion for feature extraction 78
• 4.3.4 Flow for feature extraction 78
• 4.4 Trajectory estimation 80
• 4.4.1 Recursive Bayesian state estimation with decision boundaries 80
• 4.4.2 Color-based particle filter 82
• 4.4.3 Robot localization in multi-camera networks 84
• 4.4.4 Estimating orientation by the least square method 84
• 4.5 Similarity analysis 86
• 4.6 Experimental results 91
• 4.6.1 Experimental setup 91
• 4.6.2 Results of feature extraction 94
• 4.6.3 Results of trajectory estimation 97
• 4.6.4 Results of similarity analysis 101
• 4.7 Discussion 105
• Chapter 5 Camera Calibration Using a Mobile Robot 107
• 5.1 Motivation and problem descriptions 107
• 5.2 Related work 108
• 5.2.1 Compensation of odometry errors for mobile robots 108
• 5.2.2 Camera calibration using a mobile robot 110
• 5.3 System configuration 111
• 5.4 Parameter modeling 113
• 5.4.1 Camera projection matrix 113
• 5.4.2 Differential wheeled robot kinematics 115
• 5.5 Compensation of odometry error for mobile robots 116
• 5.5.1 Definition of systematic parameters 116
• 5.5.2 Definition of non-systematic parameters 118
• 5.5.3 Estimation of systematic parameters using trajectory analysis 119
• 5.6 Estimation of camera parameters 124
• 5.6.1 Prediction 125
• 5.6.2 Update 126
• 5.7 Experimental results 128
• 5.7.1 Experimental setup 128
• 5.7.2 Results of estimated systematic parameters 130
• 5.7.3 Results of estimated camera parameters 133
• 5.8 Discussion 141
• Chapter 6 Design and Evaluation of Scalable Robotic System 143
• 6.1 System design 143
• 6.1.1 Configuration 143
• 6.1.2 Implementation 145
• 6.2 Evaluation of localization error 146
• 6.3 Evaluation of systematic error compensation 146
• 6.4 Evaluation of accuracy of calibrated cameras 160
• 6.5 Evaluation of resource sharing architecture 160
• 6.5.1 Evaluation of integration and configuration using RSA 160
• 6.5.2 Performance analysis of data transmission 162
• 6.5.3 Performance of mechanism for error handling 165
• 6.6 Discussion 167
• Chapter 7 Concluding Remarks 170
• Reference 172
• Acknowledgements 188