Xu / Qian / Wu Household Service Robotics

1. Auflage 2014
ISBN: 978-0-12-800943-7
Verlag: Elsevier Science & Techn.
Format: EPUB
Kopierschutz: 6 - ePub Watermark

E-Book, Englisch, 564 Seiten

Copyright ©2015 Zhejiang University Press, Published by Elsevier Inc. Household Service Robotics is a collection of the latest technological advances in household service robotics in five main areas: robot systems, manipulation, navigation, object recognition, and human-robot interaction. The book enables readers to understand development s and apply them to their own working areas, including: - Robotic technologies for assisted living and elderly care - Domestic cleaning automation - Household surveillance - Guiding systems for public spaces Service robotics is a highly multidisciplinary field, requiring a holistic approach. This handbook provides insights to the disciplines involved in the field as well as advanced methods and techniques that enable the scale-up of theory to actual systems. It includes coverage of functionalities such as vision systems, location control, and HCI, which are important in domestic settings. - Provides a single source collection of the latest development in domestic robotic systems and control - Covers vision systems, location control, and HCI, important in domestic settings - Focuses on algorithms for object recognition, manipulation, human-robot interaction, and navigation for household robotics

Professor, Chinese University of Hong Kong

Autoren/Hrsg.

Xu, Yangsheng

Qian, Huihuan

Wu, Xinyu

Inhaltsverzeichnis

1;Front
Cover;1
2;Household Service Robotics;4
3;Copyright;5
4;Contents;8
5;Preface;20
6;Part 1 -
Introduction;22
6.1;Chapter 1.1 - Introduction;24
6.1.1;1.1.1 Work Environments for Household Service Robots;26
6.1.2;1.1.2 Functionalities of Household Service Robots;27
6.1.3;References;35
7;Part 2 -
Service Robotic System Design;38
7.1;Chapter 2.1 - The State of the Art in Service Robotic System Design;40
7.1.1;2.1.1 Stationary Service Robotic Systems;40
7.1.2;2.1.2 Attached Mobile Service Robotic Systems;41
7.1.3;2.1.3 Mobile Household Service Robotic Systems;42
7.1.4;2.1.4 Summary of Case Studies;49
7.1.5;References;53
7.2;Chapter 2.2 - Surveillance Robot Utilizing Video and Audio Information;56
7.2.1;2.2.1 Introduction;57
7.2.2;2.2.2 System Initialization;59
7.2.3;2.2.3 Video Surveillance;63
7.2.4;2.2.4 Abnormal Audio Information Detection;70
7.2.5;2.2.5 Experimental Results Utilizing Video and Audio Information;74
7.2.6;2.2.6 Conclusions;74
7.2.7;Acknowledgments;76
7.2.8;References;76
7.3;Chapter 2.3 - Robot-Assisted Wayfinding for the Visually Impaired in Structured Indoor Environments;78
7.3.1;2.3.1 Introduction;79
7.3.2;2.3.2 An Ontology of Environments;82
7.3.3;2.3.3 RG-I: A Robotic Guide;83
7.3.4;2.3.4 Wayfinding;87
7.3.5;2.3.5 Pilot Experiments;92
7.3.6;2.3.6 Conclusions;100
7.3.7;Acknowledgments;101
7.3.8;References;101
7.4;Chapter 2.4 - Design and Implementation of a Service Robot for Elders;104
7.4.1;2.4.1 Introduction;105
7.4.2;2.4.2 Robot System;105
7.4.3;2.4.3 Human–Robot Interaction;108
7.4.4;2.4.4 Experiments;110
7.4.5;2.4.5 Conclusion;113
7.4.6;Acknowledgments;114
7.4.7;References;114
7.5;Chapter 2.5 - A Household Service Robot with a Cellphone Interface;116
7.5.1;2.5.1 Introduction;117
7.5.2;2.5.2 System Architecture;120
7.5.3;2.5.3 Grasping Algorithm;121
7.5.4;2.5.4 Solving Subproblem 1;122
7.5.5;2.5.5 Solving Subproblem 2;126
7.5.6;2.5.6 Experiments;129
7.5.7;2.5.7 Conclusion and Future Work;133
7.5.8;Acknowledgments;134
7.5.9;References;134
8;Part 3 -
Mapping and Navigation;136
8.1;Chapter 3.1 - The State of the Art in Mapping and Navigation for Household Service;138
8.1.1;3.1.1 Map Building and Localization;138
8.1.2;3.1.2 Navigation, Path Planning, and Obstacle Avoidance;143
8.1.3;3.1.3 Summary of Case Studies;145
8.1.4;References;147
8.2;Chapter 3.2 - An Error-Aware Incremental Planar Motion Estimation Method Using Paired Vertical Lines for Small Robots in Ur ...;150
8.2.1;3.2.1 Introduction;151
8.2.2;3.2.2 Related Studies;152
8.2.3;3.2.3 Problem Definition;154
8.2.4;3.2.4 Deriving a Minimum Solution with a Single Vertical Line Pair;156
8.2.5;3.2.5 Error-Aware Ego-Motion Estimation Using Multiple Vertical Line Pairs;163
8.2.6;3.2.6 Algorithms;165
8.2.7;3.2.7 Experiments;168
8.2.8;3.2.8 Conclusion and Future Work;175
8.2.9;Acknowledgments;175
8.2.10;References;176
8.3;Chapter 3.3 - Planning and Obstacle Avoidance in Mobile Robotics;180
8.3.1;3.3.1 Introduction;180
8.3.2;3.3.2 Related Work;182
8.3.3;3.3.3 Navigation Architecture;184
8.3.4;3.3.4 Roaming Trails;190
8.3.5;3.3.5 Experimental Results;193
8.3.6;3.3.6 Conclusions;202
8.3.7;References;202
8.4;Chapter 3.4 - Monocular SLAM with Undelayed Initialization for an Indoor Robot;206
8.4.1;3.4.1 Introduction;207
8.4.2;3.4.2 EKF Framework;209
8.4.3;3.4.3 Implementation of SLAM;219
8.4.4;3.4.4 Simulation and Experiment;222
8.4.5;3.4.5 Conclusions and Future Work;229
8.4.6;Appendix Supplementary Data;230
8.4.7;References;230
8.5;Chapter 3.5 - Human-Centered Robot Navigation-Towards a Harmoniously Human–Robot Coexisting Environment;232
8.5.1;3.5.1 Introduction;233
8.5.2;3.5.2 Harmonious Rules;235
8.5.3;3.5.3 Various Sensitive Fields;236
8.5.4;3.5.4 Human-Centered Sensitive Navigation System Architecture;239
8.5.5;3.5.5 Human-Centered Sensitive Navigation;240
8.5.6;3.5.6 Simulations;250
8.5.7;3.5.7 Experimental Results;255
8.5.8;3.5.8 Conclusion and Future Work;261
8.5.9;References;263
9;Part 4 -
Object Recognition;266
9.1;Chapter 4.1 - The State of the Art in Object Recognition for Household Services;268
9.1.1;4.1.1 Overview;268
9.1.2;4.1.2 Summary of Case Studies;274
9.1.3;References;277
9.2;Chapter 4.2 - A Side of Data with My Robot;280
9.2.1;4.2.1 Related Work;284
9.2.2;4.2.2 Contents and Collection Methodology;285
9.2.3;4.2.3 Contents: Robot Sensor Data;285
9.2.4;4.2.4 Annotations and Annotation Methodology;286
9.2.5;4.2.5 Annotation Methodology;288
9.2.6;4.2.6 Applications;291
9.2.7;4.2.7 Future Work;291
9.2.8;4.2.8 Related Work;293
9.2.9;4.2.9 Contents and Collection Methodology;293
9.2.10;4.2.10 Annotations and Annotation Methodology;295
9.2.11;4.2.11 Applications;295
9.2.12;4.2.12 Future Work;297
9.2.13;4.2.13 Related Work;299
9.2.14;4.2.14 Contents and Collection Methodology;299
9.2.15;4.2.15 Annotations and Annotation Methodology;301
9.2.16;4.2.16 Applications;301
9.2.17;References;305
9.3;Chapter 4.3 - Robust Recognition of Planar Mirrored Walls;308
9.3.1;4.3.1 Introduction;308
9.3.2;4.3.2 Related Work;309
9.3.3;4.3.3 Problem Definition;310
9.3.4;4.3.4 Modeling;311
9.3.5;4.3.5 Algorithm;314
9.3.6;4.3.6 Experiments;318
9.3.7;4.3.7 Conclusion and Future Work;321
9.3.8;Acknowledgments;321
9.3.9;References;321
9.4;Chapter 4.4 - Evaluation of Three Vision Based Object Perception Methods for a Mobile Robot;324
9.4.1;4.4.1 Introduction;324
9.4.2;4.4.2 Datasets and Performance Metrics;328
9.4.3;4.4.3 Lowe's SIFT;332
9.4.4;4.4.4 Vocabulary Tree Method;340
9.4.5;4.4.5 Viola–Jones Boosting;349
9.4.6;4.4.6 Discussion;354
9.4.7;4.4.7 Conclusions;355
9.4.8;Acknowledgments;356
9.4.9;References;356
10;Part 5 -
Grasping and Manipulation;360
10.1;Chapter 5.1 - The State of the Art in Grasping and Manipulation for Household Service;362
10.1.1;5.1.1 Target Detection;363
10.1.2;5.1.2 Planning;368
10.1.3;5.1.3 Control;372
10.1.4;5.1.4 Summary of Case Studies;374
10.1.5;References;374
10.2;Chapter 5.2 - A Geometric Approach to Robotic Laundry Folding;378
10.2.1;5.2.1 Introduction;379
10.2.2;5.2.2 Related Work;380
10.2.3;5.2.3 Problem Description;383
10.2.4;5.2.4 Fold Execution;386
10.2.5;5.2.5 Determining the Cloth Polygon;390
10.2.6;5.2.6 Experimental Results;397
10.2.7;5.2.7 Conclusion and Future Work;407
10.2.8;Funding;408
10.2.9;Appendix A: Proof of Theorem 1;409
10.2.10;Appendix B: Shape Models Used;410
10.2.11;Appendix C: Black Box Numerical Optimization;413
10.2.12;References;414
10.3;Chapter 5.3 - Robust Visual Servoing;418
10.3.1;5.3.1 Introduction;419
10.3.2;5.3.2 Motivation;419
10.3.3;5.3.3 Detection and Pose Estimation;421
10.3.4;5.3.4 Transportation: Coarse Visual Servoing;423
10.3.5;5.3.5 Model-Based Visual Servoing;435
10.3.6;5.3.6 Example Tasks;443
10.3.7;5.3.7 Conclusion;446
10.3.8;Acknowledgment;446
10.3.9;References;447
10.4;Chapter 5.4 - Implementation of Cognitive Controls for Robots;450
10.4.1;5.4.1 Introduction;451
10.4.2;5.4.2 Cognitive Control for Robots;451
10.4.3;5.4.3 The Working Memory System;454
10.4.4;5.4.4 The Role of CEA and FRA for Task Switching;458
10.4.5;5.4.5 Self-Motivated, Internal State-Based Action Selection Mechanism;462
10.4.6;5.4.6 Future Plans;469
10.4.7;5.4.7 Conclusions;470
10.4.8;Appendix 1. Spatial Attention and Action Selection;470
10.4.9;Appendix 2. Verbs and Adverbs for Behavior Execution;471
10.4.10;Appendix 3. Memory Contents during Work Memory Training;472
10.4.11;Appendix 4. Perception Encoding Used in FRA Experiment;473
10.4.12;Acknowledgments;474
10.4.13;References;474
11;Part 6 -
Human–Robot Interaction;476
11.1;Chapter 6.1 - The State of the Art in Human–Robot Interaction for Household Services;478
11.1.1;6.1.1 Tactile HRI Systems;478
11.1.2;6.1.2 Summary of Case Studies;483
11.1.3;References;485
11.2;Chapter 6.2 - Evaluating the Robot Personality and Verbal Behavior of Domestic Robots Using Video-Based Studies;488
11.2.1;6.2.1 Introduction;489
11.2.2;6.2.2 VHRI Methodology;490
11.2.3;6.2.3 Experiments;491
11.2.4;6.2.4 Results;498
11.2.5;6.2.5 Discussion;503
11.2.6;6.2.6 Conclusions;504
11.2.7;Acknowledgments;505
11.2.8;References;505
11.3;Chapter 6.3 - Using Socially Assistive Human–Robot Interaction to Motivate Physical Exercise for Older Adults;508
11.3.1;6.3.1 Introduction;509
11.3.2;6.3.2 Related Work;510
11.3.3;6.3.3 SAR Approach;511
11.3.4;6.3.4 Robot Exercise System;513
11.3.5;6.3.5 Motivation Study I: Praise and Relational Discourse Effects;519
11.3.6;6.3.6 Motivation Study II: User Choice and Self-Determination;527
11.3.7;6.3.7 Conclusion;534
11.3.8;Acknowledgments;534
11.3.9;References;535
11.4;Chapter 6.4 - Toward a Human–Robot Symbiotic System;538
11.4.1;6.4.1 Introduction;538
11.4.2;6.4.2 Framework for Human–Humanoid Interaction;540
11.4.3;6.4.3 Robot Memory Data Structures;547
11.4.4;6.4.4 Current Applications;550
11.4.5;6.4.5 Conclusion;552
11.4.6;Acknowledgments;554
11.4.7;References;554
12;Index;556

Leseproben

Chapter 1.1 Introduction
Huihuan Qian1,2, Xinyu Wu2,  and Yangsheng Xu1,2     1The Chinese University of Hong Kong, Hong Kong, China     2Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China Abstract
With the significant development of robotic technologies, the applications of robots have been extended broadly from their traditional industry role to military, medical field, and even daily services. Robotics has gradually become a very large industry with tremendous influence on the sustainable development of the world. Keywords
Detection; Dexterity; Motion programming; Planning; Safety; Sensing; Tele-operation Chapter Outline 1.1.1 Work Environments for Household Service Robots 5 1.1.1.1 Living Rooms/Office Rooms 5 1.1.1.2 Corridors/Narrow Paths 5 1.1.1.3 Kitchen 5 1.1.2 Functionalities of Household Service Robots 6 1.1.2.1 Human Detection and Recognition 6 1.1.2.2 Communication with Humans 6 1.1.2.3 Abnormal Event Detection 7 1.1.2.4 Floor Cleaning 7 1.1.2.5 Object Pick and Place 7 1.1.2.6 Door Opening 8 1.1.2.7 Self-Charging 8 1.1.2.8 Meal Preparation 8 1.1.2.9 Laundry Work 9 1.1.2.10 Object Carrying with Humans 9 1.1.2.11 Feeding People 10 1.1.2.12 Coaching for Elderly 10 1.1.2.13 Play with a Human 14 References 14 With the significant development of robotic technologies, the applications of robots have been extended broadly from their traditional industry role to military, medical field, and even daily services. Robotics has gradually become a very large industry with tremendous influence on the sustainable development of the world. In February 25, 2004, the International Robot Fair 2004 issued the World Robot Declaration [1] in Fukuoka, Japan. The declaration made a confident statement of the future development of robotic technology and of the numerous contributions that robots will make to all humankind. It elaborated on the expectations for the next generation robots as well as the efforts toward the creation of new markets through the next generation robots. It can be expected that the next generation robots will have a partnership with human beings, to assist human beings both physically and psychologically and to contribute to the realization of a safe and peaceful society. As Bill Gates envisioned [2], robots will be ubiquitous in every home, similar to the development trend of the computer business that began over 30 years ago. He has mentioned some researchers as the “world's best minds” trying to solve the “toughest problems of robotics, such as visual recognition, navigation and machine learning.” He also pointed out that the robotics industry was facing the challenges of nonexistence of standard operating software and standard processors, limited hardware, and the incompatibility of one programming code in another robot. Household service robots, although really appealing to everyone, still have a long way to go before entering every home and making contributions, such as has happened with personal computers. In as early as 1995, Kawaruma [3] addressed the design philosophy for service robots, emphasizing limited autonomy, which was the balance between the low autonomy in industrial robots and full autonomy in field robots. The major reasons for this were the participation of users, whom service robots are working for, and the consideration of affordable cost and system complexity. He reasonably suggested the balanced design philosophy: 1. Environmental modification: By means of low-cost environmental modifications, such as wit beacons, infrared markers, etc., the navigation problem becomes easier. 2. User–robot communication: Balance should be achieved to both take advantage of the user's intelligence and prevent tedious and exhausting tele-operations. 3. Robot intelligence: The robot should have limited autonomy in the paradigm that it has a rich set of robust, reactive behaviors, while the user can selectively activate behaviors to control. The safety mechanisms should be always operated independently and in parallel. In 1994, Kawaruma claimed [4] that the key R&D issues in service robots for the disabled and elderly include tele-operation, motion programming, planning, dexterity, sensing, and safety. Karlsson [5], in 2004, from the perspective of the robotic industry, presented three core technologies for service robots, including an object recognition system, a vision-based navigation system, and a flexible and rich software platform assisting rapid design and prototyping of robotic applications. Most of the key issues remain similar, in household service robotics as in service robots, for the disabled and elderly as presented by Kawaruma in 1994 because disabled and elderly people are one of the major groups for whom household service robots are designed to serve. Karlsson's point of view also provides a contributive subset in the key technologies for household service robots from the developer's point of view. This book provides a large number of case studies on household service robotics on a systematic level from five functionality perspectives, as shown: 1. Overall system design: presents some highlighted cases of household service robots with hardware and software structures, so that the readers can study and analyze the overall robotic system. 2. Mapping, localization, and navigation: elaborates on the problem of environment sensing, path planning, and mobility execution of household service robots. The environment may or may not need to be modified based on the different sensors resulting in different system costs. 3. Object recognition: addresses how the robots can perceive and understand the object which it should handle. This reflects the intelligence of the robotic systems. 4. Grasping and manipulation: showcases the examples on how to manipulate the objects or carry out tasks using the robot arm or other actuation structures. This reflects the dexterity of the robots. A complex task can also be planned and divided into several smaller subtasks for better efficiency. 5. Human–robot interaction: focuses on how to serve the human subject so that he/she may enjoy the convenience of and feel more natural in using the robot. 1.1.1. Work Environments for Household Service Robots
This book will illustrate the robot work scenarios in a household environment composed of living/office rooms, corridors/narrow paths, kitchens, and so forth. These environments are highly unstructured, causing more involved challenges than the traditional structured environments. Some of the figures in the next sections show these typical scenarios. 1.1.1.1. Living Rooms/Office Rooms
As a coworker, partner, or butler in a house, service robots have to adapt to the main working environment in living rooms or office rooms. The figures below illustrate some examples (Figure 1). 1.1.1.2. Corridors/Narrow Paths
The connection between different rooms, corridors, or narrow paths are also transitive environments with different features for robots to work as shown in Figure 2. 1.1.1.3. Kitchen
As one of the recent emerging application venues, service robots have more complex work in kitchens for meal preparation, dish grasping, drawer opening, etc. (Figure 3).
Figure 1 Living rooms and office rooms. (a) Office room 1 [6], (b) Office room 2 [7], (c) Office room 3 [8], and (d) Living room [9]. 1.1.2. Functionalities of Household Service Robots
Within all these scenarios, what are the required and feasible functionalities that household robots should have? This book will present the R&D achievements that are highlighted below. Due to the great passion in household service robotic research, the included robots are far from a complete set, but they provide good case studies as research references. 1.1.2.1. Human Detection and Recognition
As household service robots are aimed to serve human beings, they should be able to detect and recognize humans to ensure that they are serving the right person (Figure 4). 1.1.2.2. Communication with Humans
Communication between humans and robots is a bidirectional channel, and it should be conducted in a natural way, such as with speech, facial expressions, gestures, and so forth (Figure 5).
Figure 2 Corridors and narrow paths. (a) Corridor 1 [6], (b) Corridor 2 [10], (c) Corridor 3 [11], and (d) Corridor 4 [8]. 1.1.2.3. Abnormal Event Detection
For the security of the household environment, it is necessary for some robots to be able to classify abnormal situations from normal ones and activate an alarm as appropriate (Figure 6). 1.1.2.4. Floor Cleaning
Floor cleaning, as a routine and tedious work, may be one of the first functional tasks for commercialized...