Embed Size (px)
Citation preview
Smart MEMS a^ Sensor Systems
Sm
art M
EM
S an
d Se
nsor
Sys
tem
s D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific
.com
by T
EC
HN
ISC
HE
UN
IVE
RSI
TA
ET
on
11/1
0/14
. For
per
sona
l use
onl
y.
This page is intentionally left blank
Sm
art M
EM
S an
d Se
nsor
Sys
tem
s D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific
.com
by T
EC
HN
ISC
HE
UN
IVE
RSI
TA
ET
on
11/1
0/14
. For
per
sona
l use
onl
y.
tat IMS Elena Gaura & Robert Newman Coventry University, UK
with contributions from
Michael Kraft Southampton University, UK
Andrew Flewitt Cambridge University, UK
Davies William de Lima Monteiro Universidade Federal de Minas Gerais, Brazil
Imperial College Press
Sm
art M
EM
S an
d Se
nsor
Sys
tem
s D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific
.com
by T
EC
HN
ISC
HE
UN
IVE
RSI
TA
ET
on
11/1
0/14
. For
per
sona
l use
onl
y.
Published by
Imperial College Press 57 Shelton Street Covent Garden London WC2H 9HE
Distributed by
World Scientific Publishing Co. Pte. Ltd.
5 Toh Tuck Link, Singapore 596224
USA office: 27 Warren Street, Suite 401-402, Hackensack, NJ 07601
UK office: 57 Shelton Street, Covent Garden, London WC2H 9HE
British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library.
Cover design by John Burns.
SMART MEMS AND SENSOR SYSTEMS
Copyright © 2006 by Imperial College Press
All rights reserved. This book, or parts thereof, may not be reproduced in any form or by any means, electronic or mechanical, including photocopying, recording or any information storage and retrieval system now known or to be invented, without written permission from the Publisher.
For photocopying of material in this volume, please pay a copying fee through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA. In this case permission to photocopy is not required from the publisher.
ISBN 1-86094-493-0
Typeset by Stallion Press Email: [email protected]
Printed in Singapore by World Scientific Printers (S) Pte Ltd
Sm
art M
EM
S an
d Se
nsor
Sys
tem
s D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific
.com
by T
EC
HN
ISC
HE
UN
IVE
RSI
TA
ET
on
11/1
0/14
. For
per
sona
l use
onl
y.
Preface
This book has emerged as the result of the authors' research activity over the past six years. We started the journey as an Electronics Engineer and a Computer Scientist, both with working experience of sensing systems. The motivation for joint work was initially provided by the vision of a new world-changing technology put forward by vanguard researchers such as Kris Pister and Deborah Estrin. They foresaw systems composed of millions of MEMS sensors, collaborating in an 'intelligent' way to address many of the major problems of our age — environmental study and monitoring, pollution controls, transport safety and so on. These sensors would self-organise into networks, which in turn would be self-configuring, fully decentralised and would rely entirely on collaborative behaviour between sensors.
Like most researchers launching into a new venture, we started with simple theoretical and practical studies derived from our own sensing backgrounds and aimed at evolving the research towards full blown applications as above. Eventually it became clear that neither of our disciplines was sufficient to allow us to fully cover the scope of this research. Our vistas simply had to expand to allow us to make sensible choices as to direction and emphasis of research.
When approaching the detailed design of even quite small systems (in terms of sensor numbers), we found that the research questions posed required new science within our own respective disciplines, combined with a need for new science in other disciplines. To overcome the exposed hurdles, these findings had to be communicated to the scientists with the knowledge and skills to provide it. At this stage we became aware of how difficult it is for domain specialists to hold an overview of a topic as broad as that of large scale MEMS based sensing systems. As a simple example of the problem, a MEMS device designer may spend an enormous amount of effort optimising
v
Sm
art M
EM
S an
d Se
nsor
Sys
tem
s D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific
.com
by T
EC
HN
ISC
HE
UN
IVE
RSI
TA
ET
on
11/1
0/14
. For
per
sona
l use
onl
y.
VI Smart MEMS and Sensor Systems
a design for those last few decibels of noise margin, when, at a system level in an intelligent sensor, that same performance increase may be obtained very simply using a slightly more sophisticated signal processing algorithm. Without at least some knowledge of what is possible in each of the domains involved here (MEMS technology and smart sensor electronics), it is impossible to make sensor system design trade-offs in a sensible way. When one moves on to the possibility of sensors collaborating and operating advanced data fusion algorithms, the design choices become even more difficult and the breadth of knowledge required tenfold larger.
A study of the many new books and papers being published in the domain of intelligent MEMS systems and wireless sensor networks reveals that they mostly take a single discipline perspective. While several of these books are excellent, a researcher embarking on a top-down sensor system design venture (which we believe to be essential to achieving success in this field) is faced with assembling a library of books, each of which books assumes prior knowledge of domains that could easily be alien to our researcher.
The aim of this book is therefore to present the leading edge of this research and indeed set the research agenda in the field of MEMS based sensing. We hope to have been able to bring in a view from all of the participatory disciplines, in an integrated way. No assumption was made of a priori knowledge above the level which might be reasonably expected from, say, an electronic engineer about computer science or a computer scientist about MEMS. This has proved to be a much harder task than we initially envisaged, resolved however with the help of our distinguished contributors who have risen to the challenge of stepping out of their specialist domain and presenting deep knowledge of their own field in a way accessible to the non-specialist. Hence, we are very grateful to Dr. Andrew Flewitt for a complete and clear exposition of relevant MEMS micromachin-ing technologies; to Dr. Michael Kraft for a detailed case study on inertial sensors which reveals the merits of integrated sensor signal processing and Dr. Davies William de Lima Monteiro for a fascinating insight into the practice and potential of active optical MEMS systems.
Overall, we have tried to pay particular attention to the level of each chapter. We have not allowed ourselves to work at elementary introduction level, but have always attempted to present leading edge work in each of the three areas covered, MEMS technology, electronic system design and
Sm
art M
EM
S an
d Se
nsor
Sys
tem
s D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific
.com
by T
EC
HN
ISC
HE
UN
IVE
RSI
TA
ET
on
11/1
0/14
. For
per
sona
l use
onl
y.
Preface VII
pervasive computer science, in a digestible fashion. The central element which facilitated this treatment was the concept of the "cogent" sensor, which we developed when it became clear how overused and overloaded the terms 'smart' and 'intelligent' are, when used to describe processor integrated sensing devices. We strongly believe that it is much more useful to define things by what they do, rather than how they are constructed, when working from a top-down systems perspective. Whereas Randy Prank proposed whimsically that "A rose with a microcontroller would be a smart rose" (Understanding Smart Sensors, ppl), we would say "a rose which provides the fragrance you need would be a cogent rose". The theme of usefulness is the essence of the top-down approach.
It is widely held that smart MEMS technology will be world-changing, which presumes it will be of real use for activities which shape the world. Rather than promoting a solution waiting for a problem, we want to enable those with problems waiting for solutions to evaluate whether smart MEMS systems will rise to that challenge. If it proves to be so, we want also to enable them to begin the process of scoping that solution.
If we have achieved this goal, we believe that the book will be uniquely useful to those faced with understanding the breadth and wealth of the opportunities that combined MEMS and pervasive computing technologies offers.
Elena Gaura, Robert Newman. Coventry, 2006.
Sm
art M
EM
S an
d Se
nsor
Sys
tem
s D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific
.com
by T
EC
HN
ISC
HE
UN
IVE
RSI
TA
ET
on
11/1
0/14
. For
per
sona
l use
onl
y.
This page is intentionally left blank
Sm
art M
EM
S an
d Se
nsor
Sys
tem
s D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific
.com
by T
EC
HN
ISC
HE
UN
IVE
RSI
TA
ET
on
11/1
0/14
. For
per
sona
l use
onl
y.
CONTENTS
Preface v
Chapter 1 — Markets and Applications 1
1.1 Technology at Crossroads 1 1.2 The Present — MEMS in the News 2 1.3 The Past — Great Expectations 7 1.4 The Future — Maturity and Pervasive Applications . . . 10 1.5 Drivers for Progress 13 1.6 Progress — Device Improvement 15 1.7 Progress — Device Integration 18 1.8 Smart MEMS — The Research Agenda 23 1.9 Structure of the Book 26
Chapter 2 — Microfabrication Technologies 31
2.1 Introduction 31 2.2 Passive Components 36 2.3 Sensing Components 42 2.4 Actuating Components 49 2.5 Materials and Growth 55 2.6 Fabrication Techniques 73 2.7 Conclusions 97
Chapter 3 — Sensor Electronics 107
3.1 Introduction 107 3.2 Functions of a Sensor System 108 3.3 Analogue and Digital Design Options 131
IX
Sm
art M
EM
S an
d Se
nsor
Sys
tem
s D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific
.com
by T
EC
HN
ISC
HE
UN
IVE
RSI
TA
ET
on
11/1
0/14
. For
per
sona
l use
onl
y.
X Smart MEMS and Sensor Systems
3.4 Digital Signal Processing 137
3.5 Interface Configurations for Different Transducer Types 143
3.6 Integration 152
3.7 Design for Power Awareness 158
3.8 Conclusion 168
Chapter 4 — Sensor Signal Enhancement 173
4.1 Errors in Sensor Systems and Measurement Quality (Non-linearity, Cross-sensitivity, Offset, Parameter Drift) 174
4.2 Sensor Calibration and Compensation — Techniques and Examples 187
4.3 System Design Choices for Compensation — Closed Loop Configurations and other Designs 225
4.4 Summing up on Sensor Calibration and Compensation 226
Chapter 5 — Case Study: Control Systems for Capacitive Inertial Sensors 233
5.1 Introduction 233
5.2 Open Loop Accelerometer 235
5.3 Closed Loop Accelerometer 243
5.4 Conclusions 268
Chapter 6 — Case Study: Adaptive Optics and Smart VLSI/MEMS Systems 273
6.1 Introduction 273
6.2 Adaptive Optics and MEMS Systems 274
6.3 Operational Principles 276
6.4 Device Implementation 283
6.5 Closed-loop Adaptive Optical System 294
6.6 Conclusions and Future Trends 299
Sm
art M
EM
S an
d Se
nsor
Sys
tem
s D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific
.com
by T
EC
HN
ISC
HE
UN
IVE
RSI
TA
ET
on
11/1
0/14
. For
per
sona
l use
onl
y.
Contents xi
Chapter 7 — Artificial Intelligence Techniques for Microsensors Identification and Compensation 305
7.1 Artificial Neural Networks: What They are and How They are Used for Microsensor Control and Identification . . . . 305
7.2 Open Loop, Neural Transducer Prototype for Static/Low Frequency Applications 315
7.3 Closed-loop Neural Network Controlled Accelerometer 331
7.4 The Neural Network Non-linear Gain Controller 337
7.5 Micromachined Sensor Identification Using Neural Networks 348
7.6 Concluding Remarks 364
Chapter 8 — Smart, Intelligent and Cogent MEMS Based Sensors 369
8.1 Introduction 369
8.2 Smart, Intelligent and Cogent Sensors — What do the Terms Mean 370
8.3 What and Where is the Added Value Brought by Intelligence? 379
8.4 ANNs and MEMS 382
8.5 AI for MEMS Intelligence 390
8.6 'Cogent' Sensors — Fault Detection Case Study 403
8.7 Conclusion 412
Chapter 9 — Sensor Arrays and Networks 417
9.1 Potential of Sensor Arrays 417
9.2 Node Design 420
9.3 An Architectural History of Sensor Arrays
and Networks 425
9.4 Systems Design Issues 440
9.5 Network Technology and Topology 443
9.6 Conclusion 458
Sm
art M
EM
S an
d Se
nsor
Sys
tem
s D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific
.com
by T
EC
HN
ISC
HE
UN
IVE
RSI
TA
ET
on
11/1
0/14
. For
per
sona
l use
onl
y.
XII Smart MEMS and Sensor Systems
Chapter 10 — Wireless and Ad Hoc Sensor Networks 465
10.1 Sensor Network Applications 466 10.2 System Designers' Role 472 10.3 Design Assumptions for Ad hoc Networks 475 10.4 Distributed System Design Philosophy 477 10.5 Network Design Considerations 481 10.6 Layered Model 485 10.7 Sensor Network Operating Environments 488 10.8 Application Services 498 10.9 Proposed Sensor Support System Architecture 501 10.10 Conclusions 504
Chapter 11 — Realising the Dream — A Case Study 509
11.1 Introduction 509 11.2 The Mission 510 11.3 Initial Rough Design 513 11.4 Sensor Technology 518 11.5 Deployment 523 11.6 Operation, Control and Communication 525 11.7 Querying the Array 526 11.8 A Cogent Sensor 527 11.9 A World of Applications 530
Index 533
Sm
art M
EM
S an
d Se
nsor
Sys
tem
s D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific
.com
by T
EC
HN
ISC
HE
UN
IVE
RSI
TA
ET
on
11/1
0/14
. For
per
sona
l use
onl
y.
