The Hand eye brain System of

Intelligent Robot From Interdisciplinary

Perspective of Information Science and
Neuroscience Research on Intelligent
Manufacturing Qiao
Visit to download the full and correct content document:
https://ebookstep.com/product/the-hand-eye-brain-system-of-intelligent-robot-from-int
erdisciplinary-perspective-of-information-science-and-neuroscience-research-on-intell
igent-manufacturing-qiao/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...

The UFAW Handbook on the Care and Management of

Laboratory and Other Research Animals (UFAW Animal
Welfare) 9th Edition Huw Golledge

https://ebookstep.com/product/the-ufaw-handbook-on-the-care-and-
management-of-laboratory-and-other-research-animals-ufaw-animal-
welfare-9th-edition-huw-golledge/

Quantum Finance: Intelligent Forecast and Trading

Systems Lee

https://ebookstep.com/product/quantum-finance-intelligent-
forecast-and-trading-systems-lee/

Hieronymus Romanus Studies on Jerome and Rome on the

Occasion of the 1600th Anniversary of His Death 87 Ingo
Schaaf (Editor)

https://ebookstep.com/product/hieronymus-romanus-studies-on-
jerome-and-rome-on-the-occasion-of-the-1600th-anniversary-of-his-
death-87-ingo-schaaf-editor/

Inspirations from the Brain Master Handi Kurniawan

https://ebookstep.com/product/inspirations-from-the-brain-master-
handi-kurniawan/
The nine Eye gzi onyx A Tibetan Grammar and its
Characteristics of the Bon Tradition 1st Edition Tri
Yungdrung

https://ebookstep.com/product/the-nine-eye-gzi-onyx-a-tibetan-
grammar-and-its-characteristics-of-the-bon-tradition-1st-edition-
tri-yungdrung/

Five essays on the philosophy of nature and subject 1st

Edition Mojtaba Jafari

https://ebookstep.com/product/five-essays-on-the-philosophy-of-
nature-and-subject-1st-edition-mojtaba-jafari/

Meantone Temperaments on Lutes and Viols Publications

of the Early Music Institute Dolata

https://ebookstep.com/product/meantone-temperaments-on-lutes-and-
viols-publications-of-the-early-music-institute-dolata/

Realistic Simulation of Financial Markets Analyzing

Market Behaviors by the Third Mode of Science
Evolutionary Economics and Social Complexity Science 4
Hajime Kita
https://ebookstep.com/product/realistic-simulation-of-financial-
markets-analyzing-market-behaviors-by-the-third-mode-of-science-
evolutionary-economics-and-social-complexity-science-4-hajime-
kita/

Data Science Analytics and Applications Proceedings of

the 1st International Data Science Conference iDSC2017
1st Edition Peter Haber

https://ebookstep.com/product/data-science-analytics-and-
applications-proceedings-of-the-1st-international-data-science-
conference-idsc2017-1st-edition-peter-haber/
Research on Intelligent Manufacturing

Editors-in-Chief
Han Ding
Huazhong University of Science and Technology, Wuhan, Hubei, China

Ronglei Sun
Huazhong University of Science and Technology, Wuhan, Hubei, China

Series Editors
Kok-Meng Lee
Georgia Institute of Technology, Atlanta, GA, USA

Cheng’en Wang
School of Mechanical Engineering, Shanghai Jiao Tong University,
Shanghai, China

Yongchun Fang
College of Computer and Control Engineering, Nankai University, Tianjin,
China

Yusheng Shi
School of Materials Science and Engineering, Huazhong University of
Science and Technology, Wuhan, Hubei, China

Hong Qiao
Institute of Automation, Chinese Academy of Sciences, Beijing, China

Shudong Sun
School of Mechanical Engineering, Northwestern Polytechnical
University, Xi’an, Shaanxi, China
Zhijiang Du
State Key Laboratory of Robotics and System, Harbin Institute of
Technology, Harbin, Heilongjiang, China

Dinghua Zhang
School of Mechanical Engineering, Northwestern Polytechnical
University, Xi’an, Shaanxi, China

Xianming Zhang
School of Mechanical and Automotive Engineering, South China
University of Technology, Guangzhou, Guangdong, China

Dapeng Fan
College of Mechatronic Engineering and Automation, National University
of Defense Technology, Changsha, Hunan, China

Xinjian Gu
School of Mechanical Engineering, Zhejiang University, Hangzhou,
Zhejiang, China

Bo Tao
School of Mechanical Science and Engineering, Huazhong University of
Science and Technology, Wuhan, Hubei, China

Jianda Han
College of Artificial Intelligence, Nankai University, Tianjin, China

Yongcheng Lin
College of Mechanical and Electrical Engineering, Central South
University, Changsha, Hunan, China

Zhenhua Xiong
School of Mechanical Engineering, Shanghai Jiao Tong University,
Shanghai, China
Research on Intelligent Manufacturing (RIM) publishes the latest
developments and applications of research in intelligent manufacturing
—rapidly, informally and in high quality. It combines theory and
practice to analyse related cases in fields including but not limited to:
Intelligent design theory and technologies
Intelligent manufacturing equipment and technologies
Intelligent sensing and control technologies
Intelligent manufacturing systems and services
This book series aims to address hot technological spots and solve
challenging problems in the field of intelligent manufacturing. It brings
together scientists and engineers working in all related branches from
both East and West, under the support of national strategies like
Industry 4.0 and Made in China 2025. With its wide coverage in all
related branches, such as Industrial Internet of Things (IoT), Cloud
Computing, 3D Printing and Virtual Reality Technology, we hope this
book series can provide the researchers with a scientific platform to
exchange and share the latest findings, ideas, and advances, and to
chart the frontiers of intelligent manufacturing.
The series’ scope includes monographs, professional books and
graduate textbooks, edited volumes, and reference works intended to
support education in related areas at the graduate and post-graduate
levels.
More information about this series at http://​www.​springer.​com/​
series/​15516
Hong Qiao, Chao Ma and Rui Li

The “Hand-eye-brain” System of

Intelligent Robot
From Interdisciplinary Perspective of Information
Science and Neuroscience
1st ed. 2022
Hong Qiao
Institute of Automation, Chinese Academy of Sciences, Beijing, China

Chao Ma
School of Mechanical Engineering, University of Science and Technology
Beijing, Beijing, China

Rui Li
Institute of Automation, Chinese Academy of Sciences, Beijing, China

ISSN 2523-3386 e-ISSN 2523-3394

Research on Intelligent Manufacturing
ISBN 978-981-16-3574-8 e-ISBN 978-981-16-3575-5
https://doi.org/10.1007/978-981-16-3575-5

Jointly published with Huazhong University of Science and Technology

Press, China
The print edition is not for sale in China Mainland. Customers from
China Mainland please order the print book from: Huazhong University
of Science and Technology Press

© Huazhong University of Science and Technology Press 2022

This work is subject to copyright. All rights are solely and exclusively
licensed by the Publisher, whether the whole or part of the material is
concerned, specifically the rights of reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other
physical way, and transmission or information storage and retrieval,
electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.

The use of general descriptive names, registered names, trademarks,

service marks, etc. in this publication does not imply, even in the
absence of a specific statement, that such names are exempt from the
relevant protective laws and regulations and therefore free for general
use.

The publishers, the authors, and the editors are safe to assume that the
advice and information in this book are believed to be true and accurate
at the date of publication. Neither the publishers nor the authors or the
editors give a warranty, express or implied, with respect to the material
contained herein or for any errors or omissions that may have been
made. The publishers remain neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer

Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04
Gateway East, Singapore 189721, Singapore
Intelligent Manufacturing and Robot Theory and Technology Research
Series
Editors-in-Chief: Ding Han and Sun Ronglei
Research on the Hand-Eye-Brain of Intelligent Robot—the Interdiscipline
of Information Science and Neuroscience
Authors(s), Hong Qiao and Chao Ma and Rui Li
Huazhong University of Science and Technology Press
Preface
Robotic systems have been extensively studied during the past years
due to their various practical applications. With the rapid development
of modern industry, high-precision and compliant robotic
manipulations are significantly needed. As an important role of robotic
systems, the robots have been widely utilized in many areas. In
particular, the intelligent robot has become a hot research topic, which
is with stronger perception ability and more intelligence to make
decisions. Generally speaking, there are three main parts of the
intelligent robots: the hands, the eyes, and the brains. As a result, it is
reasonable and necessary to investigate the new approaches or
strategies for the intelligent robots. Although many pioneer works have
been reported concerning these research topics, there still remain
certain challenges and open areas.
On the other hand, the past decade has witnessed a burgeoning
development in biology researches. These remarkable results have
encouraged many bio-inspired structure or control strategies for
intelligent robots. Especially, the intelligent robots have been
investigated based on not only hardware designs but also cognition
models from the neuroscience. Meanwhile, it should be pointed out that
by exploring the intelligent robots can also help us to understand the
brain cortex better. Therefore, considering these, we will focus on
interdisciplinary aspects of biology and information science for
intelligent robots.
This monograph is concerned with the theoretical and practical
explorations on the hand-eye-brain of intelligent robot from the
interdiscipline of information science and neuroscience perspective.
More precisely, a systematic design framework is proposed for the
intelligent robots. This book can be used for academic researchers to
make further investigation for intelligent robots and may be also
suitable for engineering applications.
The contents of the book are divided into five chapters containing
several independent yet related topics, which are arranged as follows.
Chapter 1 introduces some background and the development of hand-
eye-brain for intelligent robots. Chapter 2 studies the intelligent
manipulation theory and application part for the intelligent robots.
Chapter 3 considers the intelligent perception theory and application
part for the intelligent robots. Chapter 4 addresses the intelligent
cognition theory and application part for the intelligent robots. In the
end, Chap. 7 draws the conclusions of the book and gives some future
related research directions.
Intelligent Manufacturing and Robot Theory and Technology
Research Series
Editors-in-Chief: Ding Han and Sun Ronglei
Hong Qiao
Chao Ma
Rui Li
Beijing, China
Contents
Part I Introduction
1 Background
References
2 Motivations
References
3 Structure of the Book
Part II The Hand of Robots–Intelligent Manipulation Theory and
Application of Robots
4 The Development and Challenges of Robot Manipulation
References
5 The Concept of “Attractive Region in Environment (ARIE)” and Its
Application in High-Precision Tasks with Low-Precision Systems
5.​1 General Concept of “Attractive Region in Environment”
5.​2 General Formulation of “Attractive Region in Environment”
5.​3 Extension of Attractive Region in High Dimensional Space
5.​4 Applications of “Attractive Region in Environment” in
Industrial Manufacturing
5.​5 Conclusions
References
6 The Compliance of Robotic Hands and Human-Inspired Motion
Model of Upper-Limb with Fast Response and Learning Ability
6.​1 The Combination of Compliance and Robotic Hands from
Functionality to Mechanism
6.​2 Forward Dynamics of the Upper Limb
6.​3 Movement Control Method Based on the Patterns
6.​4 Experiment Results
 6.​5 Conclusions
References
Part III The Eye of Robots–Intelligent Perception Theory and
Application of Robots
7 The Development and Challenges of Robot Perception
References
8 Learning an Intrinsic-Variable Preserving Manifold for Dynamic
Visual Tracking
8.​1 Intrinsic Variable Preserving Manifold Learning (IVPML)
8.​2 Visual-Head-Tracking Algorithm with Complicated
Backgrounds Using IVPML
8.​3 Experiment Results
8.​4 Conclusions
References
9 Explicit Nonlinear Mapping for Manifold Learning with
Neighborhood Preserving Polynomial Embedding
9.​1 Explicit Nonlinear Manifold Learning Model
9.​2 Neighborhood Preserving Polynomial Embedding (NPPE)
Algorithm
9.​3 Experiment Results
9.​4 Conclusions
References
Part IV The Brain of Robots–Intelligent Cognition Theory and
Application of Robots
10 The Development and Challenges of Robot Cognition
11 Biologically Inspired Visual Model with Memory and
Association Mechanism
 11.​1 Biological Evidences for Memory and Association
Mechanism
11.​2 Visual Model with Memory and Association Mechanism
11.​2.​1 Familiarity Discrimination
11.​3 Experiment Results
11.​4 Conclusions
References
12 Biologically Inspired Visual Model with Preliminary Cognition
and Active Attention Adjustment
12.​1 Biological Evidences for Preliminary Cognition and Active
Attention Adjustment
12.​2 Visual Model with Preliminary Cognition and Active
Attention Adjustment
12.​3 Experiment Results
12.​4 Conclusions
References
13 Biologically Inspired Visual Cognition Model with
Unsupervised Episodic and Semantic Feature Learning
13.​1 Biological Evidences for Human Visual Cognition
13.​2 Visual Model with Unsupervised Episodic and Semantic
Feature Learning
13.​3 Experiment Results
13.​4 Conclusions
References
Part V Conclusions and Future Research Directions
14 Conclusions
Part I
Introduction
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022
H. Qiao et al., The “Hand-eye-brain” System of Intelligent Robot, Research on Intelligent
Manufacturing
https://doi.org/10.1007/978-981-16-3575-5_1

1. Background
Hong Qiao1 , Chao Ma2 and Rui Li3
(1) Institute of Automation, Chinese Academy of Sciences, Beijing,
China
(2) School of Mechanical Engineering, University of Science and
Technology Beijing, Beijing, China
(3) Institute of Automation, Chinese Academy of Sciences, Beijing,
China

Hong Qiao (Corresponding author)

Email: hong.qiao@ia.ac.cn

Chao Ma
Email: cma@ustb.edu.cn

Rui Li
Email: lirui2013@ia.ac.cn

As an important part of advanced automation technologies, robots are

promising to achieve complex manipulation in various applications,
especially for industrial areas [1–3]. Generally speaking, robots are
composed of three parts: the mechanical subsystem, the sensing
subsystem and the control subsystem. With effective design, these three
subsystems can work collectively for specified tasks. However, it is
worth mentioning that with the rapid development of advanced
manufacturing industry, there are urgent need for high-precision,
compliant robotic manipulations in unstructured and flexible working
environments or human-robot interactions, which give rises to the
researches from traditional industrial robots to intelligent robots with
various application areas.
It is noted that the ability to deal with complex real-world
environment in a flexible manner is one of the most striking features of
human beings [4, 5]. As a result, one natural way to deal with the
existing challenges for intelligent robots is bio-inspired strategies and
designs from human beings. The key factors of our manipulations are
our eye, hand and brain, which govern our visual perceptions,
dexterous manipulations and cognitive decision-makings with the
environment, respectively. Particularly, in the complex and uncertain
environments, By observing the various manipulation done by human
beings, it can be found that our dexterous and successful manipulations
can express a very flexible manner. Take grasping an object for example,
among many factors that are contributing to the success of our grasping
with multiple degrees of freedom, three main aspects are especially
highlighted:
Visual Perception: acquire the enough position and shape
information of the object and the environment.
Decision making: determine how to grasp the object with certain
gesture.
Grasping motion: control the corresponding muscle to grasp the
object.
Correspondingly, the desired grasping process for robots can be
summarized as follows:
Object and environment perception: obtain feasible measurement of
the object and environment by sensory information feedback.
Grasping planning: determine the contact points on the object and
the grasping configuration.
Grasping control: motion and force control at the desired contact
points.
It can be found that by utilizing the novel bio-inspired concepts and
mimicking human behaviors, the flexibility and versatility for
intelligent robots can be potentially increased, which also helps the
researchers give an insight into the intrinsic mechanisms.
 On the other hand, it should be pointed out that although there are
considerable advantages for us in manipulations, the robots still have
certain distinguish features in high-precision tasks. Due to the
restricted capabilities of the human hand, high-precision pick up or
injection to a single cell are often implemented by robots instead of
human beings in the medical researches. However, from a practical
application perspective, achieving high-precision manipulations always
require high-precision sensing methods and structure designs for
robots. Traditionally, these requirements will correspondingly increase
the designing cost and control difficulties, which thus limit the further
applications of robots. As a consequence, it is reasonable to consider
the high-precision manipulations with relevant low-precision sensory
information for intelligent robots with bio-inspired structures and
mechanisms [6, 7].

References
1. H. Sage, M. De Mathelin, E. Ostertag, Int. J. Control 72(16), 1498 (1999)
[Crossref]
2.
M.W. Spong, S. Hutchinson, M. Vidyasagar et al., Robot Modeling and Control, vol. 3
(Wiley, New York, 2006)
3.
J.F. Engelberger, Robotics in Practice: Management and Applications of Industrial
Robots (Springer Science & Business Media, 2012)
4.
E. Pennestri, R. Stefanelli, P. Valentini, L. Vita, J. Biomech. 40(6), 1350 (2007)
[Crossref]
5.
J.M. Karl, I.Q. Whishaw, Front. Neurol. 4, 208 (2013)
[Crossref]
6.
K.Y. Goldberg, Algorithmica 10(2–4), 201 (1993)
[MathSciNet][Crossref]
7.
H. Qiao, C. Li, P. Yin, W. Wu, Z.Y. Liu, Assem. Autom. 36(1), 97 (2016)
[Crossref]
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022
H. Qiao et al., The “Hand-eye-brain” System of Intelligent Robot, Research on Intelligent
Manufacturing
https://doi.org/10.1007/978-981-16-3575-5_2

2. Motivations
Hong Qiao1 , Chao Ma2 and Rui Li3
(1) Institute of Automation, Chinese Academy of Sciences, Beijing,
China
(2) School of Mechanical Engineering, University of Science and
Technology Beijing, Beijing, China
(3) Institute of Automation, Chinese Academy of Sciences, Beijing,
China

Hong Qiao (Corresponding author)

Email: hong.qiao@ia.ac.cn

Chao Ma
Email: cma@ustb.edu.cn

Rui Li
Email: lirui2013@ia.ac.cn

During the past decade, researches on intelligent robots gains great

interest from both academic and engineering areas, and have obtained
successful achievements. Based on the above discussions, the following
issues with robot manipulation, perception and decision-making are
introduced, respectively.
Firstly, the high-precision manipulation is one of the challenging
characteristic of intelligent robots. In general, it is believed that this can
only be achieved with high-precision sensor information. It is
noteworthy that the perception ability mainly relies on the relevant
sensory precision of the robots. However, there still exist certain
limitations of high-precision sensors in hardware to date. Most sensors
of robotic hands are with low-precision, data drift or noise
characteristics, such as common visual or tactile sensors [1, 2].
Furthermore, in some applications with high-precision requirement,
certain sensors are not accurate enough, such that the obtained
information may not be always valid. Therefore, developing efficient
manipulation methods with low-precision sensors would be very
meaningful. Moreover, the existing methods for achieving compliant
motion are also important, which can be divided into three types,
namely, active compliance, passive compliance and semi-active
compliance [3–6]. Although the application of compliance is versatile,
there are two key aspects of this concept: the mechanical design with
proper physical structure and the effective manipulation with control
method and planning strategy. So far, different kinds of robotic hands,
as the main end effector of the robots, have been developed for years to
improve the structure design. One of the well-known designs is the
remote center compliance (RCC) device, which can be installed with
flexible wrists for robots [7, 8]. Meanwhile, various end effectors are
designed from simple two-fingered grippers to dexterous robotic hands
with carefully designed mechanical structures and various sensors. In
particular, researchers have been trying to introduce the functionality
of human hands to increase the compliance of the robots in terms of
bio-inspired structures, which mimicked the compliance from the
aspect of functionality.
Although remarkable achievements have been made, the
compliance of such approaches is still far from human beings.
Furthermore, another interesting observation lies in the fact that our
dexterous and successful manipulations can express a very flexible
manner with the help of a certain environment. Therefore, a natural
question arises: How do we solve these problems that are
incomprehensible for robots? Obviously, seeking and understanding the
answer to this question would greatly help in the robot designs.
Secondly, the ability of environment perception is the basis of
intelligent robots. By sensory information acquisition and processing,
the robots can well understand the surrounding environment and what
is going on. Without a sufficient environment perception, no further
manipulations or interactions can be made correspondingly. Especially,
it should be pointed out the most significant environment information
for robots is from the visual information, tactile information and
auditory information. As is well known, the sensing capability plays a
key role in the natural evolutionary process. Furthermore, we mainly
obtain the outside world information from our eyes and hands, which
provides and indicates fundamental research directions in robotic
sensory systems [9, 10]. Thanks to the rapid development of sensor and
computer vision technologies, various visual sensors with a lot of visual
processing algorithms for location, recognition and detection with
desired performances have been proposed. Nowadays, intelligent
robots with visions can be found in many practical applications, such as
industrial manufacturing, domestic services and other special robotic
areas [11, 12]. On the other hand, contrary to the visual senses, the
tactile information is not a well-defined quantity but is coarse.
Mimicking the sense of touch with rich information still remains open
and challenging [13, 14].
Following the above issues, it can be obtained that the perception
ability of robots are mainly dependent on the degrees of sensory
information acquisition. As a consequence, one primary method is
utilizing the multi-sensor information with high dimensional
information. However, at the hardware level, the perception tasks
always impose the minimum precision requirements for sensors that
may increase the cost. Furthermore, another drawback is that high
dimensional information is always difficult to extract the intrinsic
features, such that the traditional robot capacity will get considerable
limits. As a consequence, researchers started to explore, of the original
intention, if it is necessary to analyze the sensor data in a high
dimensional space, which gives rise to the so-called manifold learning
method [15, 16]. The key idea of manifold learning is that extract
intrinsic variables from high dimensional data manifolds which lie on
or close to a low dimensional manifold while keeping the data
geometric characteristics from which data samples are drawn. Up to
now, various manifold learning algorithms have been proposed with
many important applications in real-world problems [17–19].
Therefore, manifold learning based visual perception for intelligent
robots is of great importance.
 Finally, due to great advantages in the interdisciplinary aspects of
biology and information science, bio-inspired cognition strategies for
intelligent robots have drawn great attention during the past years.
With new discoveries in neuroscience, a series of computational neural
models accounting for cognition process have been proposed, which
can effective simulate the cognition process. This not only can help us
explore the deep mechanisms and structures of the brain cortex, but
also can build more invariant and discriminative computational models
for robots. For instance, the hierarchical Max-Pooling (HMAX) model
was proposed according to the hierarchical and bottom-up structure of
V1 to V4 in the ventral pathway of primate visual cortex, which could
achieve position- and scale-tolerant recognition [20, 21]. Furthermore,
with the deep understanding and unravelling of brain mechanisms for
cognitive functions, there are still rooms for the improvement of the
hierarchical cognition models for robots by introducing the novel
memory and association, preliminary cognition, motor-sensory
interactions, active attention adjustment, executive control, emotion
processing and so on [22–25]. This would also be helpful in improving
the performance of intelligent robots by simulating human movement
system.
As can be seen in the above discussions, so far, there still exists a
considerable gap between the traditional robots and intelligent robots.
Some key barriers can be how to develop the robots with artificial
intelligence during the its perception, cognition and motion process.
Attempts to bridge the existing gaps are always challenging, but the
efforts can lead to encouraging results. As a result, one of the promising
and beneficial ways regarding these issues can be the interdisciplinary
researches of information science and neuroscience with biology
evidences. This would help achieve the ultimate goal of intelligent
robots with eye, hand and brain regarding human beings as prototype.

References
1. Y. Sun, J.M. Hollerbach, S.A. Mascaro, IEEE Trans. Robot. 25(6), 1356 (2009)
[Crossref]
2.
H. Qiao, M. Wang, J. Su, S. Jia, R. Li, IEEE/ASME Trans. Mechatron. 20(5), 2311
(2015)
 [Crossref]
3.
M.T. Mason, IEEE Trans. Syst. Man Cybern. 11(6), 418 (1981)
4.
W. Wang, R.N. Loh, E.Y. Gu, Ind. Robot: Int. J. 25(1), 48 (1998)
[Crossref]
5.
R. Schiavi, A. Bicchi, F. Flacco, in ICRA’09 2009 IEEE International Conference on
Robotics and Automation (IEEE, 2009), pp. 259–264
6.
M.R. Ahmed, I. Kalaykov, in 2010 International Conference on Mechatronics and
Automation (ICMA) (IEEE, 2010), pp. 1932–1937
7.
D. Whitney, J. Rourke, J. Dyn. Syst. Meas. Control 108(3), 223 (1986)
8.
S. Lee, IEEE Trans. Autom. Sci. Eng. 2(2), 193 (2005)
[Crossref]
9.
P. Cisek, J.F. Kalaska, J. Neurophysiol. 87(2), 1149 (2002)
[Crossref]
10.
K. Fiehler, M. Burke, S. Bien, B. Rö der, F. Rö sler, Cereb. Cortex 19(1), 1 (2008)
[Crossref]
11.
F. Bonin-Font, A. Ortiz, G. Oliver, J. Intell. Rob. Syst. 53(3), 263 (2008)
[Crossref]
12.
K. Hashimoto, Adv. Robot. 17(10), 969 (2003)
[Crossref]
13.
J. Tegin, J. Wikander, Ind. Robot: Int. 32(1), 64 (2005)
[Crossref]
14.
H. Yousef, M. Boukallel, K. Althoefer, Sens. Actuat. A 167(2), 171 (2011)
[Crossref]
15.
A. Criminisi, J. Shotton, E. Konukoglu, et al., Found. Trends® Comput. Graph. Vis.
7(2–3), 81 (2012)
16.
N. Keeratipranon, F. Maire, H. Huang, Int. J. Neural Syst. 16(05), 383 (2006)
[Crossref]
17.
T. Lin, H. Zha, IEEE Trans. Pattern Anal. Mach. Intell. 30(5), 796 (2008)
[Crossref]
18.
G. Guo, Y. Fu, C.R. Dyer, T.S. Huang, IEEE Trans. Image Process. 17(7), 1178 (2008)
[MathSciNet][Crossref]
19.
H. Qiao, P. Zhang, D. Wang, B. Zhang, IEEE Trans. Cybern. 43(1), 51 (2013)
[Crossref]
20.
M. Riesenhuber, T. Poggio, Nat. Neurosci. 2(11), 1019 (1999)
[Crossref]
21.
J. Schmidhuber, Neural Netw. 61, 85 (2015)
[Crossref]
22.
C.G. Atkeson, S. Schaal, Neurocomputing 9(3), 243 (1995)
[Crossref]
23.
A. Belardinelli, F. Pirri, Cogn. Process. 7(1), 11 (2006)
[Crossref]
24.
T. Tashima, S. Saito, T. Kudo, M. Osumi, T. Shibata, Adv. Robot. 13(3), 225 (1998)
[Crossref]
25.
L. Xin, X. Lun, W. Zhi-liang, F. Dong-mei, Int. J. Adv. Rob. Syst. 10(3), 160 (2013)
[Crossref]
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022
H. Qiao et al., The “Hand-eye-brain” System of Intelligent Robot, Research on Intelligent
Manufacturing
https://doi.org/10.1007/978-981-16-3575-5_3

3. Structure of the Book

Hong Qiao1 , Chao Ma2 and Rui Li3
(1) Institute of Automation, Chinese Academy of Sciences, Beijing,
China
(2) School of Mechanical Engineering, University of Science and
Technology Beijing, Beijing, China
(3) Institute of Automation, Chinese Academy of Sciences, Beijing,
China

Hong Qiao (Corresponding author)

Email: hong.qiao@ia.ac.cn

Chao Ma
Email: cma@ustb.edu.cn

Rui Li
Email: lirui2013@ia.ac.cn

Based on the above aforementioned motivation, the structure of the

book is arranged as follows.
This chapter has introduced the system description and some
background knowledge, and also addressed the motivations of the
book.
Chapter 2 investigates the concept of “attractive region in
environment (ARIE)” and its application in high-precision tasks with
low-precision systems. The corresponding experimental simulations on
different kinds of complex manipulations show the effectiveness of the
proposed method for industrial applications. Moreover, the physical
structure and the control mechanism of human motor nervous system
to the robotic system in a tentative manner to improve the
compliance/flexibility/versatility of the robot is introduced and a novel
human-inspired motion model of upper-limb with fast response and
learning ability is proposed for robot system and control with
simulation demonstrations.
This chapter is concerned with a new intrinsic-variable preserving
manifold for dynamic visual tracking and its manifold-learning method
called IVPML, where experimental results mounted on a dynamic robot
are given to validate the effectiveness of the algorithm. Then, in order to
provide more explicit mappings for manifold learning methods, an
explicit nonlinear mapping is proposed for manifold learning, based on
the assumption that there exists a polynomial mapping between the
high-dimensional data samples and their low-dimensional
representations. Experimental results on both synthetic and real-world
data show that the proposed mapping is much more effective in
preserving the local neighborhood information and the nonlinear
geometry of the high-dimensional data samples.
Chapter 4 first considers mimicking the active memory and
association mechanism and adding the top down adjustment to the
famous biologically inspired hierarchical Max-Pooling model (HMAX
model), which can reduce the computation storage and have a good
recognition performance with a much lower memory requirement.
Then, the biologically inspired visual model with preliminary cognition
and active attention adjustment is studied to achieve a fast and robust
recognition in the retrieval and association process. The developed
model is evaluated with better performance on face databases.
Furthermore, based on the recent biological evidence, a framework to
mimic the active and dynamic learning and recognition process of the
primate visual cortex is proposed, which can achieve unsupervised
learning of episodic features and semantic features for higher level
cognition of an object.
Chapter 5 concludes the monograph by briefly summarizing the
main theoretical results presented in the book, and gives our further
investigations.
Part II
The Hand of Robots–Intelligent
Manipulation Theory and Application
of Robots

In this chapter, we consider the intelligent manipulation theory and

application of robots based on the concept of “Attractive Region in
Environment (ARIE)" and human-inspired motion model of upper-limb
with fast response and learning ability.
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022
H. Qiao et al., The “Hand-eye-brain” System of Intelligent Robot, Research on Intelligent
Manufacturing
https://doi.org/10.1007/978-981-16-3575-5_4

4. The Development and Challenges of

Robot Manipulation
Hong Qiao1 , Chao Ma2 and Rui Li3
(1) Institute of Automation, Chinese Academy of Sciences, Beijing,
China
(2) School of Mechanical Engineering, University of Science and
Technology Beijing, Beijing, China
(3) Institute of Automation, Chinese Academy of Sciences, Beijing,
China

Hong Qiao (Corresponding author)

Email: hong.qiao@ia.ac.cn

Chao Ma
Email: cma@ustb.edu.cn

Rui Li
Email: lirui2013@ia.ac.cn

Achieving high-precision manipulation with a low-precision system is a

key problem for robots, especially for manufacturing applications. So
far, sensor information feedback has been widely used to eliminate the
uncertainty of a system in many areas, which includes force sensors,
contact sensors, visual sensors and so on [1–3]. However, there are
many problems caused by sensor systems:
a sensor system increases the initial cost to the system.
 a sensor system reduces the robustness and reliability of the whole
system.
in certain cases, the states measured by sensors may be different
from that needed.
a sensor system may not be able to provide enough precision for the
specific tasks.
Furthermore, due to the high cost and maintenance difficulty of
high-precision sensors, many tedious works still need to be done
manually or with low-precision sensors. Therefore, researches on low-
precision systems that use few sensors or even no sensors to achieve
high-precision tasks are of great importance. As a result, some
remarkable works focus on eliminating the uncertainty of the system
based on limited sensor information or utilizing environment
constraints instead of additional devices to achieve high-precision
operation are reported [4, 5]. Succeeding the remarkable developments
in theory and applications, one of the efficient methods regarding these
issues is the concept of “attractive region in environment (ARIE)”,
which was inspired by the robot manipulation scenario in peg-hole
assembly operation. Successful applications to eliminate the
uncertainty of the system and achieve high-precision and high-
reliability can be found with robot assembly, grasping and localization
problems [6, 7].
On another research front line, achieving compliance is of
fundamental importance but theoretically challenging for robot
dexterous manipulation and human-robot interactions. For compliant
motions, there are generally two key factors: the mechanical robotic
hands and the corresponding compliance-based manipulation methods.
Up to now, researchers have introduced the functionality of human
hands to increase the compliance of the robotic system. Well-known
example can be found with the dexterous robotic hands with carefully
designed mechanical structures and various sensors [8–11]. However,
they are too expensive for practical use, and also the control and
maintenance is very difficult, as there are a number of joint variables
and parameters. Meanwhile, some attempts of the mechanism of
human hand movement and human motor nervous system have been
made, which may help bring in more compliance to the robots.
Therefore, it will be very meaningful and challenging to carry out the
studies on neuro-inspired hand and bio-inspired hand motion model.
Note that human movement system is a Multi-DOF, redundant,
nonlinear and complex system formed by coordinating combination of
neural system, bones, muscles and joints, which raises a complicated
problem for conventional control techniques to deal with. The motion
models simulating human movement system can be classified into
optimal control theory and habitual planning theory. Since the
computational cost of real-time optimal control for complex human
movement is tremendously large, the habitual planning tends to use
different patterns of muscle contraction to produce the same
movement rather than an online optimization process [3].

References
1. H. Qiao, B. Zhang, Proc. Inst. Mech. Eng. Part B: J. Eng. Manufact. 214(8), 657
(2000)
2.
H. Qiao, M. Wang, J. Su, S. Jia, R. Li, IEEE/ASME Trans. Mechatron. 20(5), 2311
(2015)
3.
H. Qiao, C. Li, P. Yin, W. Wu, Z.Y. Liu, Assem. Autom. 36(1), 97 (2016)
4.
K.Y. Goldberg, Algorithmica 10(2–4), 201 (1993)
5.
M. Erdmann, Int. J. Robot. Res. 14(5), 483 (1995)
6.
H. Qiao, Int. J. Prod. Res. 40(4), 975 (2002)
7.
H. Qiao, Int. J. Prod. Res. 41(14), 3159 (2003)
8.
K.B. Shimoga, Int. J. Robot. Res. 15(3), 230 (1996)
9.
A. Bicchi, IEEE Trans. Robot. Autom. 16(6), 652 (2000)
10.
M.T. Ciocarlie, P.K. Allen, Int. J. Robot. Res. 28(7), 851 (2009)
11.
E. Mattar, Robot. Auton. Syst. 61(5), 517 (2013)
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022
H. Qiao et al., The “Hand-eye-brain” System of Intelligent Robot, Research on Intelligent
Manufacturing
https://doi.org/10.1007/978-981-16-3575-5_5

5. The Concept of “Attractive Region

in Environment (ARIE)” and Its
Application in High-Precision Tasks
with Low-Precision Systems
Hong Qiao1 , Chao Ma2 and Rui Li3
(1) Institute of Automation, Chinese Academy of Sciences, Beijing,
China
(2) School of Mechanical Engineering, University of Science and
Technology Beijing, Beijing, China
(3) Institute of Automation, Chinese Academy of Sciences, Beijing,
China

Hong Qiao (Corresponding author)

Email: hong.qiao@ia.ac.cn

Chao Ma
Email: cma@ustb.edu.cn

Rui Li
Email: lirui2013@ia.ac.cn

5.1 General Concept of “Attractive Region in

Environment”
In this section, three specific applications in robotic manipulation are
analyzed to introduce the ARIE for high-precision tasks [1, 2].
Example 5.1 (Round-Peg and Round-Hole Insertion) The peg-hole
assembly operation is very difficult but necessary for automatic
manufacturing. A typical operation is round-peg and round-hole
insertion (see Fig. 5.1), whose objective is to insert a round peg into a
round hole with high precision.

Fig. 5.1 Round-peg to round-hole insertion is a typical operation in automatic

manufacturing. a is peg-rod-piston insertion for the engine assembly, and b is a shaft-
bearing insertion. In b, the gripper is grasping the shaft upper end, and the press
fitting is used for the pressing at the last procedure of assembly

As shown in Fig. 5.2, a coordinate frame is fixed on the hole to

describe the peg-hole relations. The base point of the coordinate is
defined as the center of the upper surface of the hole, the axis is
fixed on the line from to the base of the robot, the axis is
along the axis of the hole upwards, and the axis is perpendicular
to and , and satisfies the right hand rule with them.
Fig. 5.2 Coordinate frame established for the round-peg-round-hole insertion. Base
point is defined as the center of the hole’s upper surface

Then, the position of the peg in the coordinate frame built up based
on the hole can be described as

where is the constrained region formed by the edge of the hole. For

simplicity, we let , , . If the

angles between the peg and hole are fixed, then the
movement of the peg can be described in , that is,

As shown in Fig. 5.3, the motion region of the peg con-strained by

the edge of the hole has a lowest-lying point which is globally unique.
This point is corresponding to the state of three-point contact denoted
as the stable state of the assembly system, and is significant for the
insertion. This constrained region reminds us of a visualized example in
the physical space. If there is a bowl with a bean inside, and the goal is
to put the bean to the bottom of the bowl, then no matter where the
initial position of the bean is, with help of the gravity, the bean will
move to and is finally stabilized at the bottom of the bowl. For round-
peg and round-hole insertion, the peg is similar to the bean, and the
motion region of the peg formed by constraints of the hole edge is
similar to the bowl. Then, it is possible to design a state-independent
input for the peg to reduce the uncertainty of the system, just like the
gravity for the bean in the bowl.

Fig. 5.3 Configuration space of the round-peg formed by the constraints of a round-
hole
Another random document with
no related content on Scribd:
 "I'm going out for a turn," he announced presently.

Mrs. Stuart paid no attention. Once thoroughly

annoyed, she did not soon recover herself. Archie gave her
a look, then strolled out of the front door, instinctively
bending his steps in the direction of Woodbine Cottage.

He was not without hopes that Nancy might catch a

glimpse of him. Archie, of course, knew himself to be a
good-looking young fellow, especially in the tidy "second
best" suit which he donned every evening. As he passed, he
shot a side glance at the window. But no pretty blue-eyed
face was visible; and he wandered on, feeling aggrieved.
Nancy really ought to have been looking out just then!

It was a quiet spring evening, the quietness being

broken only by the subdued mixture of human and other
sounds belonging to life in a town. Absolute stillness cannot
be found there. Now and then shrill laugh, a child's cry, or a
vociferous utterance became prominent.

Suddenly a wild chorus of shouts and yells broke out,

coming nearer. Archie came to a standstill, listening. What
could it mean?

Somebody hurt! A horse running away! These thoughts

occurred to him first. He would have rushed forward to
help, but for a moment, he could not tell whence the
outcries proceeded, so full was the air of uproar.

The uproar swelled, drawing closer, shrieks mingling

with deeper-toned shouts: and all at once Archie could
distinguish words: "A MAD DOG! A MAD DOG!"

For a single instant Archie's impulse was to flee before

so gruesome a peril. But he resisted the coward-thought,
knowing that women and children might be in danger, and
went in the direction whence the volume of sound
proceeded—cautiously, and keeping a look-out.

Had he guessed who, a few seconds later, would claim

the help of his strong young arm, he would have sped to
her rescue at his utmost speed.

Not far from the neighbourhood of Woodbine Cottage

stood the large red-brick District Church, belonging to this
part of Littleburgh. It was a handsome building, free-
seated, intended mainly for working-men and their families.

Beyond the Church was the Parsonage, surrounded by a

neat garden. In this garden, near the gate, the Rev. Arthur
Wilmot stood, carefully examining a young sapling, which
seemed to have suffered a good deal from the spring
breezes.

He was a remarkably tall man, fully six feet four inches

in height, upright, vigorous, and strongly built, with a fine
thoughtful face, guiltless of whiskers or beard. It was a
genial face too, full of kindness. In age he could hardly have
passed forty, yet he had been for ten years a widower, and
the gentle girl of sixteen by his side was his only child.

"That poor little tree is done for, I am afraid," she said.

"Yes; not much hope of its recovery. But we will give it a

good dose of water, dear."

"My watering-can is nearly empty; I'll fill it again."

"No; wait, Annie. You have been often enough, my

child. Give it to me."
 Mr. Wilmot moved towards her, and at the same instant
the loud chorus of cries and yells which had startled Archie
broke upon him and Annie.

"Something must be the matter," Annie said.

Mr. Wilmot listened intently, moving closer to the gate.

Annie followed him. Suddenly words were distinguishable.
Mr. Wilmot turned to his daughter.

"Annie, go indoors at once."

He spoke in a quiet voice, but resolutely. Annie

hesitated. She had not caught the words which had reached
his ear; yet something in his look, combined with those wild
shouts, alarmed her.

"What is it?" she began to say.

"You hear me, Annie! Go indoors this instant, and close

the door. This instant: go!"

Never in his life had he spoken so sternly to his darling.

Annie turned at once to obey. She believed that he knew
something to be drawing near which he did not deem it
right or proper for her to see.

Mr. Wilmot looked after her till she had passed into the
house. Then he stepped out into the road.

Along that road, coming straight towards him, was a

black dog. The poor creature seemed exhausted, and ran
unsteadily, with drooping tail and lowered head, and bluish
foam-flecked tongue hanging from the open mouth; yet it
ran fast. A crowd of hallooing men rushed in its rear, some
little way behind.
 CHAPTER III.
THE NEIGHBOURS.

"WELL, Sue, you've made things look uncommon

comfortable, and that I must say!" Dunn observed
cheerfully, after supper. "You and Nannie must have worked
uncommon hard too."

"There's one comfort in working hard for you: that you

don't pass it all over as if it was nothing," said Susan.

"Nice sort of chap I should be, if I did! Why, it's as

clean!—And the dishes as shiny!—It's almost like the old
home kitchen, Sue!" Richard Dunn could not resist one sigh.
"And I'm sure you've made the parlour look beautiful.
Curtains up, and all!"

"Only there's no garden, father," Susie said plaintively.

"No more there is, Susie. But I'll make some good big
window-boxes, as soon as ever I can get time, and we'll fill
them quite full. And a lot of plants can stand inside the
window of the parlour too."

"Mayn't we have some mignonette?" asked Susie.

"To be sure we will. You and Nannie shall have a box of

it, all to yourselves, up at your bedroom window. Won't that
be nice? Why, Nannie,—haven't you done enough yet to-
day?" As she came in, with her neat little brown hat on.

"Dick and I are just going round into the next street,
father, for some stout thread. It'll be wanted in the
morning, and I don't want to have to go out then."

"Got lots more to do indoors, eh? But you mustn't toil

too hard, my girl."

Nancy smiled and said, "There's no fear, father."

She did not seem overdone by the day's work, standing

there in her print dress, with a pretty colour on her cheeks.

"I'm much more afraid for mother. She always will do so

much."

"Well, you're a good girl, Nannie,—always trying to save

her trouble."

"I should just hope so," Nancy answered. "I don't see
much goodness in that."

Richard Dunn took her hand, and pulled her down for a
hearty kiss. He was a very affectionate father, and he never
seemed ashamed to show his affection.

"I'm sure of one thing," he said, "and that is, that God's
blessing is on such a daughter as Nannie!"

Nancy's bright eyes grew moist with feeling. "It's nice to

hear you say so, father," she whispered; and then she stood
up. "Come, Dick, I don't want to be out late," she said.

Dick was eight years old—a rosy merry boy. He ran off
beside his sister, chattering.
 "I've had a warning about Nannie to-day," said Susan.
Dunn pricked up his ears. "Mrs. Mason came to see us,—
walked straight in through our open door, without a 'with
your leave,' or 'by your leave.' I didn't quite like it, but I do
believe she meant kindly. She lives over the way, and she
seems a nice sort of body, only such a talker. And she
begged me not to let Nancy work in a factory. She did
really!"

Susan flushed up at the recollection. Dunn burst out

laughing.

"Our Nancy work in a factory! That's a good 'un! Not if I

can help it, Sue!"

"She did really!" repeated Susan.

Dunn seemed to enjoy the idea as a capital joke. He

laughed heartily. "Our Nannie! I should think so! Not but
what they may be good girls enough,—factory girls, I mean!
But our Nannie! As if we'd ever let her!"

"Well, I'm glad you can think it funny, Richard, for I

declare I couldn't at first. But of course Mrs. Mason meant it
kindly. And oh, the things she told me! There's mothers
here, with little ones, away all day at the factories, and not
a soul at home to see after the poor mites. Even when the
men get good wages, the wives will often go out to work,
just because they don't like the dulness of home. Dulness,
indeed! I wonder what women were made for!

"And there's young girls too, our Nancy's age, getting

their ten or twelve shillings a week, and setting up
independent for themselves, and won't hear a word from
anybody, Mrs. Mason says. Not but what there's good girls
among them, as you say, good honest steady girls, I don't
doubt, who make a fight against what's wrong. But there's
too many of the other sort among them as well,—some who
won't live with their parents at all, but go off and board with
strangers. Think of that. Our Nannie's age,—and girls
younger than her too, —and the fathers and mothers
without a scrap of control over them! Why, if I had a child
like that, it would break my heart."

"Bad! Very bad!" Dunn said musingly. "It's just what

was written long ago in the Bible,—'disobedient to parents,'
you know, that's to be on the earth in the last days. Seems
to me there's a vast deal of 'disobedience to parents' in our
days."

"And a lot of blame lies with the parents themselves,"

said Susan. "If fathers and mothers are both away from
their children all day long, it passes me how the children are
to be trained. Just toss 'em out into the world, and let 'em
sink or swim. That's all the training many of them have at
all, poor little things!"

"True enough, I'm afraid," said Dunn.

"Mrs. Mason was telling me about the Gardiners next

door," resumed Susan. "The eldest girl is in one of the
factories; and her parents daren't ever say one word of
blame to her for anything, lest she should go straight off
and board with somebody else. That's a nice state of things,
isn't it? And the Handcocks on our other side,—Mrs.
Handcock is out all day long at the factory, and her husband
and children may fare as they can,—she don't care."

"I say, Sue,—seems to me Mrs. Mason is something of a

gossip," remonstrated Dunn.

Susan blushed. "Yes; only you see we've got to find out
a little about the neighbours, Richard, or we shall get
dragged in to know a lot of people that we hadn't ought to
know."

"I don't see that. Nobody can drag us into

acquaintances against our will," said Dunn. "We had best be
careful, that's all, and not get intimate with anybody in a
hurry."

"What's all that noise about?" Susan exclaimed.

The noise swelled; and both Dunn and his wife left the
kitchen, going to the front door.

People had done the same in neighbouring cottages.

"What can be the matter?" one and another was asking.

Shouts and yells seemed to sweep past near at hand,

and people could be seen running fast along a cross-road at
the end. None came past "Woodbine Cottage."

"I wish Nancy and Dick were at home," Susan said

uneasily; "I don't like this."

"Which way did they go? I'd better find them," said
Dunn.

Susan pointed out what she believed to be the right

direction. She could not feel sure, but Nancy had spoken of
shops in the next street.

Dunn started off immediately at a brisk pace, and Susan

stood in the doorway, watching and listening anxiously. The
shouts continued, but lessened somewhat, as if from
increased distance.

"Mother, do you think Nannie will be frightened?" asked

Susie, holding her mother's hand.
 "I don't know, dear. I hope not. Perhaps it wasn't really
near her,—but father will see."

"Can't think whatever in the world it's all about!" a

woman remarked, lounging listlessly at the door of "Rose
Cottage," to Susan's left. She was a careworn slatternly
unhappy-looking woman: and the smudged faces and dirty
frocks of the children by her side were unpleasant to see.

"Mother!" wailed one dismally, "Mother—Jacky's pinched

me!"

"I didn't," screamed Jacky's shrill tones.

"He did, mother!"

"Have done with your quarrelling, will you! I declare,

you're all the plague of my life!" And two or three sounding
slaps were administered round with great impartiality,
producing a burst of shrieking sobs.

"Be quiet, now,—will you! If you don't hold your

tongues, I'll slap you both again, that I will!" cried Mrs.
Gardiner. "There's no peace in life with your squabbling."

The children certainly did not hold their tongues, and

Mrs. Gardiner lugged two or three of them indoors,
bestowing a shake upon each by the way.

"Mother, was that little girl naughty?" asked Susie's

wondering tones. "The boy really did pinch her, and he
made ugly faces too—I saw him."

"I'm afraid the mother didn't take much trouble to find

out who was wrong," Susan said softly, and not very wisely,
perhaps. There was no need to call the child's attention to
Mrs. Gardiner's shortcomings.
 Susie's small voice sounded clearly in answer, "Dick
doesn't pinch me. I'm so glad he doesn't. But he did scratch
me once, mother,—don't you know? And father made him
go straight to bed, and Dick was so sorry after. He went and
got me two big bull's-eyes. And you didn't slap me, did you?
'Cause it wasn't my fault?"

"Well done, little 'un!" laughed a hearty voice on the

other side.

Susan glanced towards the big broad-shouldered

working-man, seated in the doorway of "Myrtle Cottage,"
smoking. She rather liked his look, and she liked too the
way in which a puny little boy had confidently climbed upon
"father's knee." That spoke well for the man. Susan was
much less attracted by the hard features and gaudy cap of
the woman who stood behind him, drawn out by the noise.
But she felt very anxious, and she could not refrain from
asking, "I suppose you can't tell me what's the matter?"

"No, missis, I can't," the man answered civilly. His name

was Handcock, as Susan guessed rightly; and he alone, of
all the men within sight, had not started off to see what was
up. "Shouldn't wonder if it's a lot of boys chasing a cat."

But the voices were those of men, not boys; and almost
immediately a cry came down the street: "Mad dog! Mad
dog!"

CHAPTER IV.
NANCY AND THE DOG.
 CHILDREN were bundled promiscuously within cottage
doors, and mothers followed them. Handcock stood up at
last. He seemed an easy-going sort of individual—indolent
perhaps, and not readily startled or flurried by passing
events.

"O Nannie!" Susan had said faintly, at the first instant.

Then she sent Susie indoors, and stood close to it herself,
waiting.

The crowd did not come that way, neither did the dog.
People began presently to breathe more freely, and
Handcock returned to say, "There's no fear. Shouldn't
wonder if it's all a scare about nothing!" But he had not
gone farther than the end of the street to inquire.

Suddenly Richard Dunn hurried up, Dick trotting by his

side.

"Nannie not here?" Dunn said breathlessly.

"No, she hasn't come. O Richard!"

"It's a scare about the mad dog, I suppose?" said

Handcock.

"A scare! No. It's true!" said Dunn hoarsely. "And

Nannie not here!"

"Tell me quick,—haven't you seen her?" asked Susan,

holding on to his arm, for he seemed about to start off
without another word.

"No. She left something in the shop, and Dicky ran back
for it. And the crowd came between—dog and all. Dick
wasn't allowed to leave the shop. Nobody seems to know
anything about Nannie. Let me go, Susan."

Susan made no effort to detain her husband.

Handcock said, "I'll come with you, neighbour;" but his

movements were too slow for the distracted father. Dunn
was gone.

Somebody spoke to Susan, but she made no answer.

She could not stand, for her legs shook under her, as if with
the palsy, and she crept inside the cottage, and sat down.
Both children came close.

"Mother, I wouldn't have left Nannie if I had known the

dog was coming," said Dicky. "And I wanted to go to her,
but the woman held me tight, and locked the door. I
thought Nannie would be so frightened. Do you think the
dog has hurt her, mother?"

Susan shook her head, and moaned. She could not shed
a tear, and her parched tongue refused to speak.

Somebody's face was put in at the door.

"Here she comes, Mrs. Dunn! All right!" a voice said.

The owner of the voice withdrew, and Nancy entered.

She walked with a hasty faltering step, and her face was
perfectly white,—lips, cheeks, and all, as colourless as
chalk. Close behind followed a tall good-looking young man.

Susan stood up, and took bold of Nancy with a tight

grasp. She wanted to ask, "Are you hurt?" but no words
came.
 "Hadn't you better both sit down?" asked the young
man. "The dog didn't touch her, Mrs. Dunn,—thank God, he
didn't! She's only had a fright."

"Mother, he saved me!" Nannie said. "I should have

been bitten but for him!" And Nancy burst into a flood of
tears.

* * * * * *

Yes, Archie Stuart had saved Nancy—at his own risk—

from one of the most terrible perils which can well beset a
man or woman.

When the thought of helping somebody had occurred to

him, he could not, of course, guess who that somebody
might be.

Nancy had failed to find exactly what she wanted at the

nearest shop in the next street: so she had gone farther.
Halfway home she found that a small pattern for the colour
of the thread, which she knew her mother would need, had
been left behind.

Dick offered to run back for it, and Nancy, feeling tired,
consented. She promised to wait for Dick a quiet sort of
back lane, which had a high wall on either side, broken only
by one five-barred gate, leading into a yard, and locked.
Nancy did not even notice the gate.

She had strolled but a few paces, when a burst of yells

filled the air.

What could be the cause? Nancy felt a little afraid. She

wished too that she had not sent Dick off alone.
 Scant time remained for thought. The shouts drew
nearer, and the warning cry,—"A mad dog!" reached Nancy's
ears distinctly. Almost at the same instant, a black dog
appeared at the farthest end of the lane, running straight
towards her, dropping foam from his open jaws and hanging
tongue.

Nancy staggered against the wall, sick with horror. Men

followed behind, but none were near enough to succour her.
She could see no outlet—no means of escape. Her limbs
seemed paralysed with the shock, so that she could not
even run.

"O God, help me! Oh, help!" That cry went up from
Nancy's heart, as the sense of her peril grew upon her. It
was not the vague despairing cry of one who has never
thought of God till danger or death threaten. Nancy had
known God from very infancy as her great and loving Father
in heaven,—as One to whom, in the Name of Christ, she
might always have full and instant access. She had a
childlike trust in His great might. It was the instinct of her
heart to cry to Him in need.

Nancy did not see it, but help was at hand. Her dilated
eyes, fixed upon the poor mad creature rushing towards
her, could look in no other direction.

Archie Stuart had entered the nearer end of the lane,

behind Nancy, just before the dog appeared.

He did not recognise her at first as the girl whose pretty

and gentle face had taken his fancy. But without recognising
Nancy Dunn, he did recognise a woman in danger,—and,
happily, that was enough.

What to do? was the question. Alone he could have

climbed the wall without difficulty, for he was an active
young fellow. But with Nancy! That made all the difference.

The gate lay a few paces beyond Nancy, between her

and the dog. The best hope lay there, if it could be reached
in time. Flight seemed hopeless.

"Don't be frightened!" a voice said at her side. "Here,—

this way!"

Nancy clutched her deliverer's arm with one sharp cry;

and before she knew what was happening, he had dragged
her to the gate.

Some men in Archie's place might have thrown

themselves to the top first, and pulled up the girl after. But
somehow Archie could not do that. All the manliness of his
nature revolted from putting himself in safety while a
woman stood below within reach of the dog.

Without losing an instant, he lifted Nancy in his strong

arms, raising her rapidly till her feet were on the top bar
except one.

"Hold fast!" he said shortly, sharply, breathing hard in

his excitement. He did not look to see how close the
dreaded creature had come; but he felt that in four or five
seconds more—

"Oh—come!" gasped Nancy.

"Hold on hard! You're safe!"

Archie loosened his grasp of her, clutched the top of the

gate, and swung himself up, with a desperate effort, barely
in time. As his feet rose, the open jaws followed, but the
passing snap was fruitless. One spring, and Archie stood
upon the topmost bar, drawing the terrified girl higher, lest
the dog should leap. Already, however, the creature had
gone on, and the crowd of shouting men swept after,
keeping, it must be confessed, at a respectful distance.

Then Archie sprang to the ground, and lifted down the

trembling girl. In the white frightened face, he now
recognised Dunn's pretty daughter.

"It's all right; the dog won't come back," Archie said
encouragingly. "It's a mercy I came. But we'll get out of this
lane as soon as we can. It's an awkward place to be caught
in. Anywhere else you might have got inside a door."

"Please,—now," Nancy managed to say, and though

shaking still, she was able to walk. At the end of the lane
she paused suddenly.

"Dick—where can Dick be?"

"Is he your brother?"

"Yes. He went back to the shop. What can have become

of him?" asked Nancy, in distress.

"They'll have kept him there, you may be sure of that.

I'll just get you home, and then go to see after him. Don't
be frightened," Archie urged once more.

"It was so dreadful!" Nancy whispered. "I can't thank

you. Father will."

"Isn't his name Dunn?" asked Archie, putting a very

unnecessary question. "I walked home with him to-day. My
name's Stuart, and he used to know my father; but I dare
say he didn't think of mentioning me?"
 Nancy was unable to give the answer which Archie
desired. She could remember nothing at that moment
except the mad dog, her own past peril, and her fears about
Dick.

A few minutes brought them to Woodbine Cottage,

where Dick was found to have arrived before them. And
while Nancy was still sobbing, and Susan looking stupefied,
and Archie standing by, half proud, yet embarrassed, Dunn
walked in.

He had been told outside of Nancy's return, "looking as

white as paper," one woman said, "and scarce able to drag
herself along." Nobody seemed to be sure whether she had
entirely escaped injury. Dunn heart was filled with a terrible
foreboding.

"Nannie!" he said hoarsely, "Is anything wrong?" Then

he turned to Archie. "Tell me the worst," he muttered.

"She's not hurt," Archie answered quickly. "She's not

hurt, indeed; only frightened. The dog didn't touch her; I
give you my word for it, he didn't. She'll be all right
presently."

"Nannie, he speaks the truth?" pleaded poor Richard

Dunn, hardly able to believe what he heard.

Nancy looked up at her father, with blue wet eyes, her

chest heaving still, and her lips quivering. "Yes, father, it's
quite true," she sobbed. "I'm not hurt. But I should have
been—if—if—O father, he saved me! I should have been
bitten but for him!"

Dunn grasped Archie's hand, and turned his face away,

unable to express what he felt. And none who know what
true manliness is will think one whit the worse of that
strong working-man for the big tears upon his cheeks,
drawn forth by the thought of his child's peril.

"Tell me about it," he said huskily, after a pause.

Mrs. Mason had just then walked in, but nobody took
any notice of her. She said nothing, only lifting her hands
and eyes with gestures of mute astonishment, and then
bringing a mug of water to Nancy, which helped to check
her sobbing.

Archie was by no means unwilling to "tell all about it."

He did not wish to exalt unduly his own courage, but of
course he knew that he had been courageous, and he felt
much gratification in having saved somebody—especially
pretty Nancy Dunn—from a great danger.

So he narrated what had occurred, neither underrating

nor overrating his own share in the matter. Nancy tearfully
corroborated his tale, even to the snap of the dog's jaws
after Archie's heel, which she had distinctly heard.

"Words can't thank you, my lad. If ever there's anything

I can do for you,—if it were with my last shilling,—I'd do it!"
Dunn said fervently. And then, taking off his cap, which he
had hitherto forgotten to remove, Dunn added, in a
reverent voice:

"Thank God for His great mercy!"

Susan had neither stirred nor spoken hitherto, and as

yet no one had particularly noticed her stunned look. But
now Mrs. Mason went to her side with the mug of water,
and made her drink some. And Susan at length looked up,
saying in a hollow voice, "I shall never trust her out of my
sight again."
 "There! There! There! Poor dear! It's given her such a
turn, you see," Mrs. Mason remarked soothingly, patting
Susan on the back, and looking round at everybody
present.

"I shall never trust her out of my sight again," repeated

Susan.

"Sue, I wouldn't feel that," Dunn said.

"She might have been bitten! And I should never have

forgiven myself—sending her out alone!"

Mrs. Mason patted Susan again, and Dunn came nearer.

"Sue, my dear, I wonder who takes best care of our

Nannie," he said,—"you and me, or God? If you'd been with
her in the lane, I wonder what you could have done for her?
Wasn't it better, God sending the help in time? I wouldn't
feel that, Sue. It seems to me, we ought to be feeling that
there's no need ever to fear again—she's been so cared for,
just when you and me couldn't get at her, or help her."

Susan sighed heavily. "Yes, you're right," she said. "But

if Nannie had been bitten—if we'd the thought of that before
us!"

"I suppose God could bear us through anything—even

that!" Dunn said, with an effort. "Only He hasn't called us to
it. He has taken care of our Nannie for us. I do think we
ought to be full of nothing but thankfulness, instead of
talking about what might have been. I do think it," repeated
Dunn.

"I ought to go home," Archie said unwillingly. "Mother 'll

be anxious, if she hears about the dog."
 "Well, we won't keep you, lad," said Dunn. "But you'll
come again? You can't come too often. I shall be able to
thank you better another day. Sue and me feel all turned
upside-down-like this evening."

Archie went off briskly, straight home first, though

determined later to inquire what was become of the dog. He
found Mrs. Stuart in a very troubled state. Some rumours of
peril to Archie had reached her ears; and she was working
herself up into a fever of fright. Archie's appearance and his
bright face were a great relief.

He told her something of what had happened, but not

more than was necessary. He rightly judged that Mrs.
Stuart would not welcome any event which should tend to
intimacy with the Dunns.

Other people, however, lavishly filled up gaps in Archie's

tale; and before nightfall, Mrs. Stuart knew as much of
Archie's doings as he knew himself.

CHAPTER V.
A BRAVE DEED.

ANNIE WILMOT, passing quickly indoors at her father's

bidding, did not catch the ominous words which had
reached his ears. And believing that he wished her not to
see the cause of the uproar, she went dutifully to the