Download the full version of the textbook now at textbookfull.

com

Enabling the Internet of Things From

Integrated Circuits to Integrated Systems 1st
Edition Massimo Alioto (Eds.)

https://textbookfull.com/product/enabling-the-
internet-of-things-from-integrated-circuits-to-
integrated-systems-1st-edition-massimo-alioto-eds/

Explore and download more textbook at https://textbookfull.com

 Recommended digital products (PDF, EPUB, MOBI) that
you can download immediately if you are interested.

From Internet of Things to Smart Cities: Enabling

Technologies 1st Edition Hongjian Sun

https://textbookfull.com/product/from-internet-of-things-to-smart-
cities-enabling-technologies-1st-edition-hongjian-sun/

textbookfull.com

TFET Integrated Circuits: From Perspective Towards Reality

Navneet Gupta

https://textbookfull.com/product/tfet-integrated-circuits-from-
perspective-towards-reality-navneet-gupta/

textbookfull.com

The Internet of Things and Big Data Analytics: Integrated

Platforms and Industry Use Cases 1st Edition Pethuru Raj
(Editor)
https://textbookfull.com/product/the-internet-of-things-and-big-data-
analytics-integrated-platforms-and-industry-use-cases-1st-edition-
pethuru-raj-editor/
textbookfull.com

Campfire Chaos (Hood River Hoodlums #3) 1st Edition K

Webster [Webster

https://textbookfull.com/product/campfire-chaos-hood-river-
hoodlums-3-1st-edition-k-webster-webster/

textbookfull.com
Interaction-induced Electric Properties of van der Waals
Complexes 1st Edition Victor N. Cherepanov

https://textbookfull.com/product/interaction-induced-electric-
properties-of-van-der-waals-complexes-1st-edition-victor-n-cherepanov/

textbookfull.com

Tandem Mass Spectrometry of Lipids Molecular Analysis of

Complex Lipids Robert C Murphy

https://textbookfull.com/product/tandem-mass-spectrometry-of-lipids-
molecular-analysis-of-complex-lipids-robert-c-murphy/

textbookfull.com

Getting away with murder the true story of the Emmett Till
case Crowe

https://textbookfull.com/product/getting-away-with-murder-the-true-
story-of-the-emmett-till-case-crowe/

textbookfull.com

Sugarcane Biorefinery Technology and Perspectives 1st

Edition Fernando Santos (Editor)

https://textbookfull.com/product/sugarcane-biorefinery-technology-and-
perspectives-1st-edition-fernando-santos-editor/

textbookfull.com

Oxford textbook of neuroscience and anaesthesiology First

Edition. Edition Engelhard

https://textbookfull.com/product/oxford-textbook-of-neuroscience-and-
anaesthesiology-first-edition-edition-engelhard/

textbookfull.com
Crazy in Loathe An Enemies to Lovers Romantic Comedy All
My Jingle Ladies Book 4 1st Edition Melissa Williams

https://textbookfull.com/product/crazy-in-loathe-an-enemies-to-lovers-
romantic-comedy-all-my-jingle-ladies-book-4-1st-edition-melissa-
williams/
textbookfull.com
Massimo Alioto Editor

Enabling the
Internet of
Things
From Integrated Circuits to Integrated Systems
Enabling the Internet of Things
Massimo Alioto
Editor

Enabling the Internet

of Things
From Integrated Circuits
to Integrated Systems
Editor
Massimo Alioto
Department of Electrical & Computer Engineering
National University of Singapore
Singapore, Singapore

ISBN 978-3-319-51480-2 ISBN 978-3-319-51482-6 (eBook)

DOI 10.1007/978-3-319-51482-6

Library of Congress Control Number: 2016963262

# Springer International Publishing AG 2017

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or
part of the material is concerned, specifically the rights of translation, 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 publisher, 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 publisher 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.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
To Maria Daniela and Marco, and now also Marina.
My deepest merriment begins with M. And them, indeed.

To my family, including my beloved nephews and niece:

Rachele, Davide, Gaetano and Francesco.

And to everyone who has inspired my curiosity and love for life,
including Giusi, Rodolfo, Alfio, Dora, Gaetano, Annamaria
and Santina. And thankfully many, many others.
Preface

Decades of exponential improvements in integrated circuit manufacturing

and design have been spurred by (and given rise to, in a virtuous circle)
relentless reduction in the cost per transistor, as well as many other interest-
ing consequences. The cost reduction keeps following a well-proven learning
curve, which includes Moore’s law as a corollary, and will continue in spite
of the end of this law. According to the Bell’s law, such technology trend has
led to inexorable shrinking of electronic systems, which are now approaching
the sub-centimeter scale and below. At the same time, the Koomey’s and
Gene’s laws promise the reduction in the energy consumption by another two
orders of magnitude. Those trends promise the possibility of integrated
electronic systems that are very inexpensive, small, and extremely low
power. In other words, we will increasingly see systems that are pervasive
in space and long-lived in time. At the same time, Metcalfe’s law (or
Sarnoff’s law for the least optimistic) traces the fast-growing value of
connectivity, thanks to the rapidly increasing number of connected users,
and, more in general, connected objects.
At the same time, several megatrends are demanding more pervasive and
continuous sensing, as well as sensemaking and transfer of physical data.
Accelerated urbanization and increasing worldwide population requires sus-
tainable usage and sharing of resources, as well as more livable and smarter
environments at all scales (from home to city). Pervasive sensing and
sensemaking are also being required by assistive and proactive technologies
(e.g., robotics, decision support) that increasingly relieve humans from
routine tasks, repetitive labor, and recently data-driven decision-making.
The sharing economy is demanding the ability to spatially track, to physically
monitor and manage objects, to encourage responsible usage, and to charge
users by the actual usage. Well-being and other human factors are being
modeled and monitored to create healthy environments where humans can be
happy and productive. Geosocialization and participatory sensing are pro-
gressively involving objects other than individuals or as support to human
activities. Three-dimensional remote physical interaction with reality
provides sensory feedback, thus demanding ubiquitous sensing to enable
this ability on a wider scale and on a finer granularity.
The push and the pull effect of the above technological trends and
applications is converging on and creating a virtuous circle that we now

vii
viii Preface

call the “Internet of Things” (IoT). The IoT can evidently create a huge value
and bring unprecedented benefits to the society. To set this on a trend
perspective, we can extrapolate Hick’s law to artificial intelligence and
cloud computing: more physical data will enable us to take more automated
decisions with an effort that is only logarithmic in the space of decision
choice. The IoT is ultimately a powerful enabler to share on a larger scale,
make technology more human centric and real time, and decouple socioeco-
nomic progress from intensive use of resources. And, interestingly, IoT
silicon technology becomes so small that the user is immersed in it (there
is no more “user experience,” in a sense), with interesting implications in
terms of market and perceived value.
In spite of the daily IoT-related claims in the chip design community, the
tiny sensing nodes of the IoT at its edge (the “IoT nodes”) are still in their
technological infancy. Several challenges need to be tackled, such as energy
efficiency and related lifetime, cost, security, and interoperability, among
others. Such challenges need to be tackled in a holistic manner, developing
both an understanding of the different parts of IoT nodes and an insight into
the big picture and the strong linkage to applications and related
requirements.
To the best of our knowledge, this is the first book on integrated circuit
and system design for the Internet of Things. This book develops in both the
“vertical” and the “horizontal” dimension. Vertically, it provides a compre-
hensive view on the challenges and the solutions to successfully design chips
for IoT nodes as systems (from circuits to packages), a broad analysis of how
chip design needs to evolve to meet those challenges, and a fresh perspective
grounded on historical and recent trends. Horizontally, the book covers in
one place the very diverse domain-specific expertise of the subareas involved
in the design of IoT nodes, which was previously scattered across a large
number of talks, journals, and conferences.
This book provides a design-centric perspective, providing an understand-
ing of what the IoT really means from a design point of view. Typical
specifications of commercial IoT nodes are discussed, and constraints
imposed by IoT applications are translated into design constraints that chip
designers are used to deal with. Design guidelines to meet them are system-
atically discussed in every chapter.
This book started in the form of talks at various venues, such as VLSI
Symposium, HotChips, and ISCAS, where I had very interesting
conversations with several other speakers. Those talks were motivated by
the lack of a cohesive and detailed source of accessible knowledge on the
design of IoT nodes. The idea to write this book came exactly from those
conversations, which later continued throughout the interaction with chapter
authors. They really made this book possible, providing their deep insights
and invaluable expertise. I deeply thank all outstanding researchers and
designers who contributed to the chapters of this book, sharing their expertise
in an accessible and concise manner for the benefit of our community.
Preface ix

This book is structured as follows. Chapter 1 describes the big picture in

view of technological trends, an overview of the challenges ahead and the
possibilities that research has recently opened, and some link to the econom-
ics of the IoT and social megatrends. Chapter 2 provides a system-level
perspective of IoT nodes. Then, Chaps. 3–7 cover the design of digital
subsystems of IoT nodes, from architectures to circuits, and memories in
CMOS and other emerging technologies. Chapter 8 is about hardware-level
security techniques, whereas Chap. 9 focuses on System-on-Chip design
methodologies. Power management and energy harvesting are covered in
Chaps. 10 and 11. Analog interfaces and analog–digital converters are
discussed in Chaps. 12 and 13. Short-range radios are discussed in Chap.
14. Batteries as further essential component of IoT nodes are the focus of
Chap. 15. Packaging is the topic of Chap. 16. Finally, Chaps. 17 and 18
describe two system integration examples, exemplifying the design
techniques introduced in the previous chapters. As a common thread, all
chapters include a final section on perspectives and trends, which provides a
glance into the future, and a good starting point for further research and
advances.
There are many ways to use this book. In particular, it can serve as a
reference to practicing engineers working in the broad area of integrated
circuit/system design of IoT nodes, in view of the wide and detailed coverage
of state-of-the-art solutions for IoT and the fresh perspective on the future of
such technologies. The book is also very well suited for undergraduate,
graduate, and postgraduate students, thanks to the rigorous and lean coverage
of topics and selected references.

Singapore Massimo Alioto

December 2016
Contents

1 IoT: Bird’s Eye View, Megatrends and Perspectives . . . . . 1

Massimo Alioto
2 IoT Nodes: System-Level View . . . . . . . . . . . . . . . . . . . . . . 47
Pascal Urard and Mališa Vučinić
3 Ultra-Low-Power Digital Architectures for the Internet
of Things . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Davide Rossi, Igor Loi, Antonio Pullini, and Luca Benini
4 Near-Threshold Digital Circuits for Nearly-Minimum
Energy Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Massimo Alioto
5 Energy Efficient Volatile Memory Circuits
for the IoT Era . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
Jaydeep P. Kulkarni, James W. Tschanz, and Vivek K. De
6 On-Chip Non-volatile Memory for Ultra-Low Power
Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
Meng-Fan Chang
7 On-Chip Non-volatile STT-MRAM for
Zero-Standby Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
Xuanyao Fong and Kaushik Roy
8 Security Down to the Hardware Level . . . . . . . . . . . . . . . . 247
Anastacia Alvarez and Massimo Alioto
9 Design Methodologies for IoT Systems on a Chip . . . . . . . 271
David Flynn, James Myers, and Seng Toh
10 Power Management Circuit Design for IoT Nodes . . . . . . . 287
D. Brian Ma and Yan Lu
11 Energy Harvesting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
Ying-Khai Teh and Philip K.T. Mok
12 Ultra-Low Power Analog Interfaces for IoT . . . . . . . . . . . 343
Jerald Yoo

xi
xii Contents

13 Ultra-Low Power Analog-Digital Converters for IoT . . . . . 361

Pieter Harpe
14 Circuit Techniques for IoT-Enabling Short-Range
ULP Radios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
Pui-In Mak, Zhicheng Lin, and Rui Paulo Martins
15 Battery Technologies for IoT . . . . . . . . . . . . . . . . . . . . . . . 409
Jeff Sather
16 System Packaging and Assembly in IoT Nodes . . . . . . . . . . 441
You Qian and Chengkuo Lee
17 An IPv6 Energy-Harvested WSN Demonstrator
Compatible with Indoor Applications . . . . . . . . . . . . . . . . 483
Pascal Urard, Liviu Varga, Mališa Vučinić,
and Roberto Guizzetti
18 Ferro-Electric RAM Based Microcontrollers:
Ultra-Low Power Intelligence for the Internet
of Things . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 503
Sudhanshu Khanna, Mark Jung, Michael Zwerg,
and Steven Bartling
Visit https://textbookfull.com
now to explore a rich
collection of eBooks, textbook
and enjoy exciting offers!
About the Editor

Massimo Alioto (M’01–SM’07-F’16) was

born in Brescia, Italy, in 1972. He received
the Laurea (M.Sc.) degree in Electronics
Engineering and the Ph.D. degree in Elec-
trical Engineering from the University of
Catania (Italy) in 1997 and 2001, and a
Bachelor of Music in Jazz Studies from
the Conservatory of Music of Bologna
in 2007.
He is currently an Associate Professor at
the Department of Electrical and Computer
Engineering, National University of Singapore, where he leads the Green IC
group and is the Director of the Integrated Circuits and Embedded Systems
area. Previously, he was Associate Professor at the Department of Informa-
tion Engineering of the University of Siena. In 2013 he was also Visiting
Scientist at Intel Labs—CRL (Oregon) to work on ultra-scalable microarch-
itectures. In 2011–2012, he was Visiting Professor at the University of
Michigan, Ann Arbor, investigating on active techniques for resiliency in
near-threshold processors, energy-quality scalable VLSI design, and self-
powered circuits. In 2009–2011, he was Visiting Professor at BWRC—
University of California, Berkeley, investigating on next-generation ultra-
low power circuits and wireless nodes. In the summer of 2007, he was a
Visiting Professor at EPFL—Lausanne (Switzerland).
He has authored or co-authored more than 220 publications in journals
(80+, mostly IEEE Transactions) and conference proceedings. One of them is
the second most downloaded TCAS-I paper in 2013. He is co-author of three
books, Enabling the Internet of Things—From Integrated Circuits to
Integrated System (Springer, 2017), Flip-Flop Design in Nanometer
CMOS—From High Speed to Low Energy (Springer, 2015) and Model and
Design of Bipolar and MOS Current-Mode Logic: CML, ECL and SCL
Digital Circuits (Springer, 2005). His primary research interests include
ultra-low power VLSI circuits, self-powered and wireless nodes, near-
threshold circuits for green computing, energy-quality scalable VLSI
circuits, hardware-level security, circuits for on-chip learning, and circuit
techniques for emerging technologies.

xiii
xiv About the Editor

Prof. Alioto was a member of the HiPEAC Network of Excellence

(EU) and the MuSyC FCRP Center (US). In 2010–2012 he was the Chair
of the “VLSI Systems and Applications” Technical Committee of the
IEEE Circuits and Systems Society, for which he was also Distinguished
Lecturer in 2009–2010 and member of the DLP Coordinating Committee in
2011–2012. He is also member of the Board of Governors of the IEEE
Circuits and Systems Society (2015–2017). In the last 5 years, he has given
50+ invited talks in top universities and leading semiconductor companies.
He currently serves as Associate Editor-in-Chief of the IEEE Transactions on
VLSI Systems, and served as Guest Editor of various journal special issues
(e.g., IEEE TCAS-I issue on Internet of Things in 2017, IEEE TCAS-II issue
on green computing in 2012). He also serves or has served as Associate
Editor of a number of journals, such as IEEE Transactions on VLSI Systems,
ACM Transactions on Design Automation of Electronic Systems, IEEE
Transactions on CAS—part I and part II, Microelectronics Journal, and
others. He serves or has served as panelist for several funding agencies and
research programs in the USA and Europe. He was Technical Program Chair
(ICECS, PRIME, VARI, NEWCAS, ICM, SOCC) and Track Chair in a
number of conferences (ICCD, ISCAS, ICECS, VLSI-SoC, APCCAS, ICM).
Prof. Alioto is an IEEE Fellow.
 IoT: Bird’s Eye View, Megatrends
and Perspectives 1
Massimo Alioto

This chapter opens the book and provides a sum- its pervasive networking, cloud and related
mary of the challenges and the opportunities that advances (e.g., big data), the physical world
are offered by the Internet of Things (IoT), with through distributed sensing and people’s
emphasis on the aspects that are relevant to activities, in the unprecedented form of mostly
integrated circuit and system design from circuits real-time fine-grain and aggregated data from the
to packaging for IoT nodes. The chapter is knowledge coming from environments, goods,
organized along a chronological perspective, resources, tools, infrastructures, among the others.
first reviewing technology historical trends So far, the IoT has been defined in several
beyond mere Moore’s law, and summarizing different ways, and its meaning has become so
recent past achievements and capabilities that broad that it oftentimes includes any object on
are making the IoT possible. Then, present earth that is connected to the Internet, such as
challenges are described, as pathway to connected cars, drones, smartphones, smart
up-coming advances and developments in the appliances, industrial tools, and so on. Under
design of IoT nodes. Finally, mega-trends are such generic definition based on pure Internet
examined to unearth clues on longer-term evolu- connectivity, the IoT has been already realized
tion of the IoT and the implications on integrated as the number of computing devices connected to
system design. the Internet surpassed the worldwide population
back in 2008–2009 (Evans et al. 2011).
This book focuses on the IoT as pervasive,
1.1 The Internet of Things: Context unobtrusive, systematic and coordinated intro-
and Overview duction of sense-, compute-, communication-
ability and sensemaking of physical data in a
The concept of the IoT seems to first appear in very large number of objects on earth. This is
Kevin Ashton in a presentation delivered at enabled by the introduction of extremely
Procter & Gamble in 1999 (Ashton 2009), which miniaturized integrated systems (“IoT nodes”)
was then described as a large-scale network of with very long lifetime (e.g., decades) that are
smart RFIDs. On a broad perspective, the IoT autonomous in many respects, from functional-
lies at the intersection of the Internet realm with ity, to energy, to the way they interact with the
physical world and the network infrastructure.
M. Alioto (*) From this perspective, the IoT pushes such
National University of Singapore, Singapore 117583, capabilities beyond personal devices (e.g.,
Singapore smartphones), embedding them in everyday
e-mail: [email protected]

# Springer International Publishing AG 2017 1

M. Alioto (ed.), Enabling the Internet of Things, DOI 10.1007/978-3-319-51482-6_1
2 M. Alioto

objects and living environments. This book predict a somewhat slower growth (Nordrum
addresses the challenges involved in the creation 2016). The IoT market size is expected to have a
of IoT nodes in the form of integrated circuits, global economic impact of 2.5–11.1 T$ by
covering the different areas involved in this pro- 2022–2025 (Dobbs et al. 2015; Jankowski et al.
cess including architecture, circuit building 2014).
blocks, design methodologies, packaging and As shown in the simplified architecture in
system demonstrations. Being the IoT an exten- Fig. 1.1, the IoT is structured into three tiers of
sive topic, the scope of this book purposely devices. At the bottom, IoT nodes perform sensing
excludes the challenges related to the integration and interact with the physical world. To assure
of IoT nodes into a cohesive and scalable net- scalability and ubiquitous network access, gate-
work comprising inter-operable and heteroge- ways and concentrators collect, protect (under
neous nodes, and related communication users’ control) and route data from several and
protocol and software layers. physically proximal IoT nodes, and route it to
A commonly agreed target of the IoT is to servers. The latter perform data aggregation and
expand the number of connected devices per per- knowledge extraction, and deliver physically-
son to the order of a thousand, thus reaching an enhanced cloud services. Some additional inter-
unprecedented scale of trillions of connected mediate levels of aggregation might be needed,
devices (Gaudin 2015). The number of depending on the amount of data generated, the
connected devices is expected to grow to 30–50 area covered by a sub-network, and the density of
billion devices by 2020, with an expected market IoT nodes, among the others. For example,
CAGR growth of 15–35% (Markets and Markets; concentrators might actually be a sub-set of the
https://newsroom.cisco.com/press-release-content network below an Internet hub/gateway, which is
?type¼press-release&articleId¼1771211; Ericss here omitted as this would be simply part of the
on Mobility Report; http://www.gartner.com/ existing Internet infrastructure.
newsroom/id/3165317; Greenough and Camhi The hardware requirements of the devices in
2015; Worldwide Internet of Things Forecast the three tiers in Fig. 1.1 are very different, by
2015; TechNavio 2015; Machina Research 2015; virtue of their significantly different number and
Bauer et al. 2014; Jankowski et al. 2014; Dobbs level of pervasiveness. The number of IoT nodes
et al. 2015; Digital Universe of Opportunities is expected to be approximately two orders of
2014), [IoT Analytics, Oct. 2014], (http://www. magnitude larger than the number of con-
postscapes.com/internet-of-things-market-size/). centrators, which in turn is plausibly higher than
Some forecasts question such fast growth and the number of server blades by another two orders

# DEVICES WORLDWIDE
size
cost/item USERS compute- communication- sense-
power ability ability ability status
100 Millions
illions 100-1,000
meterss GFLOPs well deployed
CLOUD 10,000 $
1,000-10,000 keeps expanding
1,000 W
10 Billions
ions 100-1,000
INTERNET GATEWAYS/ 10 cm MFLOPS well deployed
OF THINGS CONCENTRATORS 10+ $ keeps expanding
1 -10 W
1 Trillion
on 1 -100 commercial nodes
1 -10 mm MFLOPS still far from IoT
IoT NODES
1$ target
000 mW
0.1-1,000

PHYSICAL
WORLD

Fig. 1.1 A simplified architecture of the IoT

1 IoT: Bird’s Eye View, Megatrends and Perspectives 3

of magnitude. To be embeddable in objects and following sections. As a result, IoT technologies

the living environment, the form factor of IoT currently tend to substantial fragmentation, pos-
nodes is expected to be in the scale of millimeters, ing a fundamental challenge in terms of economy
which is at least an order of magnitude smaller of scale and interoperability, which adds to the
than concentrators, whose size is expectedly in the expected fragmentation due to lack of
same order as today’s wireless routers (10-cm standardization in this early phase of its
range). The form factor of server blades is another development.
order of magnitude larger. The cost target for IoT Internet cloud services and wireless networks
node is widely accepted to be in the 1-dollar range will be greatly affected by the expansion of the
(Ricker et al. 2016), as might be expected by IoT, due to the large number of connected nodes.
observing that an average customer in the con- The IoT is indeed being responsible for data del-
sumer electronics market would likely spend as uge issues that impact the network traffic, and the
much as a top-of-the-line smartphone to populate power associated with wireless communications.
their home and objects with 1000 IoT nodes. Regarding the data deluge, currently only 1% of
Concentrators are allowed to have a larger cost enterprise data is being used to generate valuable
in view of their lower number, expectedly by an knowledge, and is mostly utilized for alarms or
order of magnitude at least, considering their real-time control (Dobbs et al. 2015). Such poor
non-trivial computational and wireless bandwidth data utilization for useful purposes and value
requirements. In turn, cloud servers clearly entail creation will further worsen due to the volume
a larger cost by at least two orders of magnitude. increase determined by the IoT, as the worldwide
Similarly, concentrators are expected to deliver at data is expected to grow by 4 in 2015–2020
least two orders of magnitude more compute (Digital Universe of Opportunities 2014;
power compared to IoT nodes, which can Jankowski et al. 2014; https://newsroom.cisco.
typically have very limited (e.g., sub-Mega com/press-release-content?type¼press-release&
Operations per Second—MOPS) or moderate articleId¼1771211). By 2020, data generated by
computational capability (100 MOPS). Cloud IoT devices will account for 10% of the world’s
servers are certainly required to have a much data (Digital Universe of Opportunities 2014)
larger computational power compared to (i.e., approximately 44 zettabytes). Hence, the
concentrators, by at least two orders of magnitude. IoT will demand better data utilization as well
Due to their large number and ubiquity, IoT as pre-selection and filtering of valuable data to
nodes need to be untethered and hence their be processed and stored in the cloud. Regarding
power budget is very small, and is as low as the volume of wirelessly transmitted data, in
sub-μW for miniaturized systems powered by 2020 the IoT is expected to generate 1000
energy harvesters. Due to their larger size and more data than in 2015 (Digital Universe of
lower density, concentrators are expected to be Opportunities 2014), with an overall power con-
mostly tethered, and hence their power can be sumption that would become comparable to the
much larger (e.g., in the order of Watts). A server expected total worldwide energy production of
blade dissipates a power that is two orders of 25 PWh (Callewaert 2016). Accordingly, the
magnitude larger. wireless power consumption in the IoT needs to
IoT nodes have design requirements that are be substantially reduced for sustainability
markedly different from existing Internet- reasons, which adds to the issues raised by the
connected devices (e.g., networked computers tight power limitations of IoT nodes (see later). In
and smartphones), as they aim at facilitating addition, the large number of IoT nodes requires
convergence of several tasks onto a single plat- an acceleration in the transition from the 32-bit
form (Jankowski et al. 2014). Instead, IoT nodes IPv4 Internet protocol suitable for 4  109 dif-
need to pursue hardware specialization and appli- ferent addresses, to the 128-bit IPv6 protocol that
cation specificity, mostly for the very stringent can handle up to 1038 addresses, with some chal-
power requirements, as discussed in the lenge imposed by the different 64–96 bit length
4 M. Alioto

of RFID identifiers (Atzori et al. 2010). Finally, buildings and nations, toys, worksites, smart
the Internet as we know it today was mostly infrastructures, energy, lifestyle/entertainment,
designed for non-real-time sharing of documents among the others.
and data, with resiliency being the main concern As opposed to previous technological waves
(Greenemeier et al. 2013). Due to the generation in the semiconductor history, the IoT is the first
of large amounts of real-time data, the IoT pushes one that is so pervasive that it becomes invisible
the Internet towards its limit and hence needs to to the users, with several implications on the
be structured in a more decentralized manner to value capturing in the semiconductor industry.
assure sustainable scalability. For example, only 5–10% of the IoT technology
The above issues related to the IoT data del- spending is expected to fuel the semiconductor
uge are drastically mitigated by moving intelli- industry market (Dobbs et al. 2015), whereas
gence from the cloud to the concentrators and more value (15–20% each) will be captured by
most importantly to the IoT nodes in Fig. 1.1, software and integration services. Plausibly,
i.e. making the IoT nodes “smarter” most of the value of the IoT will come from the
(or “cognitive”, if intelligence means ability to data aggregation and the real-time response
detect and classify patterns) than they are today. (or actuation) of cloud services, as well as the
Indeed, pre-processing in the IoT nodes and more demand prediction for new proactive services
distributed intelligence reduce the data volume, and tasks that no longer need us to “push a
as only partially aggregated data needs to be sent button” (or click a mouse) to be executed. To
over the network, as opposed to raw data. capture more value from the large market volume
and by delivering integration services (e.g., from
IoT nodes to software for data aggregation and
1.2 Brief Review of IoT sensemaking), semiconductor companies will
Applications likely become more vertically integrated through
acquisitions, close partnerships and industrial
1.2.1 Considerations on the IoT consortia. As further benefit, this trend will also
Market Volume favor IoT node inter-operability and
standardization.
The IoT as a whole is inherently a general-
purpose technology, similarly to computers and
mobile devices in the past decades. Like any 1.2.2 Summary of Current
other general-purpose technology, it can boost and Prospective Applications
true productivity and create a value that is sub- of the IoT
stantially higher than its market size, as it can
serve as catalyst for bigger change (Brynjolfsson The IoT is a very fragmented application sce-
and Hitt 1998). Indeed, the IoT can further nario (Vermesan and Friess 2014), and
improve efficiency, economy of scale, ability to encompasses a wide range of applications, some
react to and predict demand in capex, labor and of which are summarized in the following.
energy. Also, the IoT is expected to enable better In the agriculture sector, the IoT infrastructure
coordination and usage monitoring of buildings, can monitor the quality, the actual usage and the
machinery, manufacturing processes, factories, availability of resources, for better and predictive
supply chain and resources. The IoT will impact management (e.g., irrigation) and storage (e.g.,
a very wide diversity of applications, from agri- avoid waste of feed and fertilizing). Monitoring
culture to consumer products, automotive, the environmental conditions permits to support
healthcare, retail, manufacturing and supply the growth of animals and plants (e.g., aquacul-
chains (e.g., Industry 4.0), telecommunications, ture), optimally time the next course of action,
logistics, public sector, financial, transportation and ultimately assure quality (e.g., wine) and
and shipping, smart environments from homes to raise the efficiency in the production process.
1 IoT: Bird’s Eye View, Megatrends and Perspectives 5

In automotive, the IoT enables the monitoring objects for personal care and hygiene can be
of the state of a vehicle down to its critical used to remind of regular but infrequent care
components, from initial shipping to usage, to activities based on dentist’s suggestions shared
assess their correct utilization (e.g., detecting in the cloud, and motivate positive behavior in
bumps, vibration) and maintenance (e.g., open- children. Smart clothing can remind of periodic
ing of containers, wearing parts). Based on actual cleaning based on actual usage. Smart toys can
usage, predictive maintenance can be performed be selectively enabled only upon the occurrence
to lengthen the vehicle lifetime, and lower the of desired conditions to create positive habits
upkeep. Such capabilities enabled by the IoT are (e.g., only at certain times or lighting conditions),
also very useful in fleet management and car and prevent danger by disabling them under the
sharing services. Also, distributed sensing and presence of others (e.g., toddlers). Smart jewelry
global sensemaking enables traffic control can be used to unobtrusively track activity, mea-
through differentiated and personalized road sure exposition to solar light and other environ-
pricing to encourage virtuous behavior and pri- mental conditions, and make emergency calls.
oritize tasks for commercial (e.g., car pooling Energy management at different scales can be
with multiple passengers sharing cost) and pri- made more effective by the IoT. At the city scale,
vate vehicles (e.g., fast delivery for critical the smart grid offers several opportunities to
goods), through virtual/dynamic city area leverage the sensing and sensemaking
boundaries. capabilities of the IoT to optimize the energy
In public transportation, the occupancy and usage across many users, a better coordinated
utilization can be monitored to assure an ade- usage and planning of alternative energy sources,
quate quality of service, detect potential danger ultimately reducing the overall energy and the
(e.g., potential collision between vehicles and currently large gap between the peak and the
pedestrians), and predict short term demand average consumption.
based on crowd monitoring in strategic locations. Health care is another important application
On the road side, excessive congestion and pol- area in which the IoT promises to fundamentally
lution can be managed with real-time demand- contribute to. As few examples, the miniaturiza-
response schemes where the road pricing is tion and long lifetime of IoT nodes provides an
dynamically adjusted through real-time obser- unobtrusive mean to constantly monitor vital
vations and utilization prediction, based on signs and other related parameters (e.g., behav-
previous history and real-time data in strategic ioral) and develop deeper understanding of the
locations. Also, the transportation of dangerous patient’s health evolution. In addition, the avail-
goods and the circulation of slow (or frequently ability of big data from a large number of patients
stationary) vehicles can be optimally coordinated offers an unprecedented opportunity to explore
with the ordinary traffic to minimize their nega- correlations, build models and tools for predic-
tive impact. Again, the IoT offers unprecedented tive diagnosis, early treatment and make drug
opportunities to share resources efficiently, while discovery more efficient and effective. Similar
preserving their running condition. considerations hold for the elderly and the
Consumer electronics substantially benefits disabled, as constant non-obtrusive monitoring
from the IoT, as its pervasiveness permits to allows for better and highly responsive/predic-
track smart goods (e.g., positioning systems for tive care, while preserving individual’s indepen-
object retrieval) and detect their exposure to dency and offloading hospitals. Remote
anomaly conditions (e.g., overheating, physical supervision also enhances the ability to share
shocks). IoT sensing can signal spatial professionals across a larger number of
co-presence of objects and specific people (e.g., individuals and patients, thus driving the care
kids) to signal potential danger, or to recommend cost down.
activities to complete when all necessary objects Industrial processes and logistics can also
tools are available in the same space. Smart highly benefit from the IoT, as it can enable
6 M. Alioto

ubiquitous sensing of operating conditions, real- Smart homes promise to automatically order
time tracking of semifinished products, detection supplies before exhaustion and upon the avail-
of events that slow down the process throughput ability of online offers, as a major step forward
and potential safety issues. The data generated in compared to the today’s Amazon Dash wireless
the production line can be intelligently shared button that simply orders goods online when the
with the quality assurance process and across button is pushed (Amazon Dash). At the same
different sites, to raise the yield and reduce time, waste management will be made more effi-
cost. On the warehouse side, product location cient by sensing the actual demand, and pricing
and storage conditions can be tracked for more based on actual consumption habits, thus encour-
efficient product delivery and distribution. Simi- aging virtuous behavior. The IoT can also make
larly, sharing real-time and historical data on residential compost recycling easier and
parts with the procurement process makes automated, through the monitoring of humidity
restocking more efficient, and reduces the inven- and temperature trends.
tory cost through more strategic purchasing Smart buildings can leverage the IoT to be
strategies. The presence of IoT within machines more adaptive to the actual demand and needs
can enable early and self-diagnosis, predictive of the occupants, while ensuring the highest
(rather than reactive or pre-scheduled) mainte- safety and comfort standards. Indeed, air quality
nance, pre-emptive vendor support to prevent and thermal/acoustic/visual comfort can be mon-
known failures. Again, the IoT enables better itored and controlled for the first time with a
economy of scale, efficiency and makes pro- granularity that goes down to the single room,
cesses leaner. with obvious advantages in terms of comfort
In the area of retail, smart malls can provide assurance and energy cost. Beyond normal build-
real-time shopping recommendations, matching ing operation, the real-time capability of the IoT
available offers with individual customers, dis- enables the ability to respond to critical events
card products for potential customers with (e.g., fire) quickly, minimizing the human and
allergy issues and provide other personalized material losses in case of emergencies.
services (e.g., for customer fidelization). The Through the IoT, smart cities can manage
tight coupling between the individual and collec- resources more efficiently, be made much more
tive customers’ behavior, the store setting and resilient to temporary malfunctions and disasters,
the warehouse permits to streamline the inven- and encourage virtuous behavior. Smart and
tory management, offer better shopping experi- weather-adapting lighting, water/gas leakage
ence, dynamically adjust in-store display based monitoring, smart parking with dynamic pricing
on the predicted demand, and cut inventory costs. and area allocation, no physical boundaries and
Through the IoT, smart homes can manage automated parking advice are just a few
utilities more efficiently by controlling individ- examples of how to use the IoT to solve today’s
ual appliances based on actual utilization and urban challenges. Ubiquitous vision can enable
needs, and purchasing electricity when cost is an unprecedented level of safety and security,
lowest within the day in demand-response energy detecting potential danger and provide crucial
pricing schemes. Unprecedented levels of secu- information on crowd behavior and citizens’
rity (e.g., perimeter access control) are achiev- needs (e.g., for adaptive and predictive transpor-
able thanks to the pervasiveness of IoT nodes and tation management, real-time digital signage
sensemaking ability. Occupant recognition recommendations to prevent immediate danger).
permits to adjust lighting, sound, air condition- Other than enabling ubiquitous and augmented
ing/heating based on individual preferences. This surveillance, vision in IoT offers physical aug-
can be done in a predictive manner, so that mentation to social media and recommendation
occupants do not need to “push any button”, systems (e.g., venue recommendation based on
leveraging the fine-grain knowledge of crowdedness, and crow sentiment), and human
occupants’ habits and the ability of the cloud to activity monitoring to achieve better match
generalize and extract trends and predictions. between demand and supply of services
1 IoT: Bird’s Eye View, Megatrends and Perspectives 7

dynamically. Similarly, distributed audition

permits to develop situational awareness, build
real-time noise urban maps to mitigate noise form
pollution at critical times, and localize noise factor
events for safety assurance. Smart irrigation of access to
security
green spaces and parks is another sub-area where power
the IoT has potential to make an impact. Smart
tourism promises to give tourists the ability to cost lifetime
have an immediate understanding of the city,
such as availability, crowdedness or quietness
of different places to receive dynamic wireless
always
recommendations on tours that adapt to their connected

disposition, other than already available factual sensing &

processing
information on places. Waste management can
be made more efficient and priced fairly as
discussed before, while detecting potentially
dangerous and inappropriate waste that would Fig. 1.2 Summary of IoT node requirements
need to be disposed with different procedure
(again, encouraging virtuous behavior).
required capabilities of IoT nodes and user
Smart infrastructures will also benefit from
requirements, as summarized in Fig. 1.2 and
the IoT in terms of safety (e.g., structural moni-
discussed in the following.
toring of bridges) and security (e.g., automated
identification of unattended bags and suspicious
behavior). Distributed IoT nodes will enable
gesture-based natural human-infrastructure inter- 1.3.1 Physical Constraints
face, where users do express their preferences
anywhere and any time, being constantly Physical constraints of IoT are dictated by size
observed, instead of pushing buttons on an elec- considerations and the necessity to untether IoT
tronic booth or controllers (e.g., thermostat). This nodes and avoid any maintenance (e.g., battery
also introduces the new capability to average out replacement), as dictated by their large number.
requests and preferences from multiple users, Regarding the form factor, IoT nodes need to be
thanks to the distributed nature of such human- sufficiently small to make the deployment of IoT
infrastructure interface. nodes non-intrusive, with a typical volume rang-
In the areas of wildlife and nature preserva- ing from cubic millimeters to hundreds of mm3.
tion, the IoT can monitor both the activity and the Being untethered, IoT nodes need to be energy
living conditions of wildlife, as well as the qual- autonomous and rely on a battery and/or an
ity of available natural resources (e.g., water), energy harvester as energy source (Alioto 2012).
their level of pollution, forest fire detection, In purely battery-powered nodes, the average
earthquake early detection, counteraction of ille- IoT node power Pavg needs to be small enough to
gal activities against wildlife. achieve the desired lifetime tlifetime ¼
Ebattery =Pavg , for a given battery energy capacity
Ebattery. Figure 1.3 shows the lifetime of an IoT
1.3 Requirements of IoT Nodes system versus its average power consumption,
assuming optimistically that the battery self-
The distinctive features of IoT nodes nodes are leakage and ageing are negligible. From this
defined by the requirements imposed by IoT figure, smartphone or button cell batteries assure
applications in terms of physical constraints, a reasonably long lifetime of a decade (or more)
type of interaction with the external world, for Pavg in the order of few hundreds of nWs.
8 M. Alioto

1E+12
1E+10
10 years
1E+8
1 year
lifeme (s)

1E+6 1 week
1E+4 1 hour
1E+2 1 minute
1E+0
1E-2
nW mW mW
1E-9 1E-8 1E-7 1E-6 1E-5 1E-4 1E-3 1E-2 1E-1 1E+0
power (W)
smartwatch battery button cell battery thin-film battery

type cost capacity volume energy density

GH43-03992A 30$ 300 mAh 2,400 mm3 0.12 mAh/mm3
LR44 <1$ 150 mAh (non- 500 mm3 0.28 mAh/mm3
rechargeable)
Cymbet CBC005 0.2$ 5 mAh 0.7 mm3 6.5 mAh/mm3

Fig. 1.3 Lifetime vs. average power consumption for different batteries

Larger Pavg mandate the addition of an energy to deliver the peak power, if the former does not
harvester, whose size is generally proportional to have adequate instantaneous power capability, as
Pavg. Figure 1.4 shows the harvester size required dictated by its size.
for a given Pavg for various energy sources. From As opposed to purely battery-powered systems,
this figure, millimeter-sized photovoltaic energy harvested IoT nodes can operate nearly-
(indoor), thermo-electric (on-body patch) and perpetually, as long as the harvester power
airflow (indoor) harvesters can indefinitely sus- exceeds Pavg (i.e., the harvester size is large
tain Pavg in the order of μWs (Alioto 2015). Tens enough), and can hence indefinitely sustain the
of μWs are sustainable under more abundant power required by the IoT node. On the other
energy sources, such as photovoltaic (outdoor), hand, an increase in the targeted lifetime tlifetime
thermo-electric (industrial machines) and body ¼ Ebattery =Pavg for a given Pavg requires the adop-
vibration (e.g., walking) harvesting (Alioto tion of proportionally larger batteries. Hence,
2015). GSM radio-frequency energy harvesting energy harvesters are invariably more compact
can instead sustain only tens to very few than batteries for long enough lifetime targets.
hundreds of nWs. From Fig. 1.3, printed and Figure 1.4 shows the breakeven lifetime at which
solid-state batteries (see Chap. 15) enable harvester and battery have the same size, assum-
aggressive miniaturization at the cost of much ing a battery with energy density equal to typical
shorter lifetime, whose extension requires the alkaline button cell batteries (e.g., LR44). From
addition of an energy harvester in all practical this figure, harvesting is always more compact for
cases. As a third energy source option, the battery all practical lifetime targets under abundant
can be suppressed altogether by pairing the energy sources, such as photovoltaic (outdoor),
energy harvester with a small energy source thermo-electric (industrial machines) and body
(e.g., off-chip supercapacitor, on-chip capacitor) vibration (e.g., walking) (Alioto 2015). On the
Visit https://textbookfull.com
now to explore a rich
collection of eBooks, textbook
and enjoy exciting offers!
1 IoT: Bird’s Eye View, Megatrends and Perspectives 9

Fig. 1.4 Lifetime 1000

vs. average power
consumption for different
batteries
100

harvester size (mm)

10

IoT node
1
size target

0.1
0.1 1 10 100 1000 10000
avg power Pavg (mW)
photovoltaic (indoor) photovoltaic (outoor)
thermal (human) thermal (industrial)
vibration/motion (human) air flow (indoor)
electromagnetic (GSM)

other hand, harvesters become more compact than other living environments and infrastructures,
batteries for targeted lifetimes of 2–3 years and this translates into a lifetime of several decades.
longer, and hence in most of IoT applications. Industrial applications, transportation and
GSM radio-frequency harvesters are instead shipping might require a shorter lifetime,
always larger sized than the battery counterpart. although still in the order of a decade. The life-
Regardless of the specific lifetime target and time requirement can be further relaxed in other
energy source architecture, the volume of IoT applications such as retail, worksites, lifestyle/
nodes is certainly dominated by off-chip entertainment. Hence, the above considerations
components, and in particular by the energy on the power budget of IoT nodes apply almost
source, as the antenna can be made very thin. In unmodified in a very wide range of applications.
other words, the size of IoT nodes is essentially Meeting power budgets of few μWs or below is
set by their power consumption, which hence feasible only if the IoT node actively performs
become a very stringent and crucial requirement tasks (e.g., sensing, processing) only infrequently.
in any IoT node design. In other words, power needs to be aggressively
reduced by duty cycling the IoT node operation,
alternating active tasks and long sleep periods as
depicted in Fig. 1.5, with periodicity set by the
1.3.2 Interaction with the External
wake-up cycle Twkup. From an architectural stand-
World
point, this means that IoT nodes are organized into
an always-on (ALWON) sub-system that manages
With reference to Fig. 1.2, the interaction of IoT
the periodicity of the wake-up cycle and stores
nodes with the external world needs to last at
information across active tasks, and a duty-cycled
least the lifespan of the object/environment they
(DCYC) sub-system that periodically performs the
are embedded in, as battery replacement is not an
active task (Alioto 2012). Hence, the average IoT
option due to the large number or the inaccessi-
node power can be written as the sum of the
bility of nodes. When deployed in buildings or
10 M. Alioto

Fig. 1.5 Lifetime active mode active mode

vs. average power
consumption for different
batteries
sleep mode sleep mode

Twkup

ALWON power PALWON and the DCYC energy designed resolution (and hence higher cost and
EDCYC (Alioto 2012): power). From Fig. 1.6, most of IoT applications
require a minimum resolution that is below
EDCYC
Pavg ¼ PALWON þ : ð1:1Þ 12 bits, and 8 bits are sufficient for a rather
T wkup wide range of practical cases. On the secondary
From Eq. (1.1), the power reduction can be y axis, the figure also reports the energy per
reduced by reducing the power (energy) of the conversion, assuming an energy per conversion
always-on (duty-cycled) sub-system. In other step of 30 fJ, which is relatively optimistic espe-
words, nearly-minimum power design needs to cially for larger resolutions (Murmann 1997).
be pursued in the always-on sub-system, while The datarate range of the above sensors is plot-
nearly-minimum energy design is the objective ted in Fig. 1.7. This figure shows that most of the
in the duty-cycled one (see Chap. 4). Of course, sensors require only thousands of bits per second
larger Twkup and hence more infrequent active when operating continuously, whereas tasks
operation mitigates power, although Twkup is related to vision and audio processing need orders
upper bounded by the application, depending on of magnitude higher datarates (up to 10 Mbps in
how frequently data needs to be updated. Such the case of compressed VGA video streaming).
system-level tradeoffs are discussed in Chap. 2, From the above considerations, the specifications
whereas approaches to further reduce the leakage of IoT node sensing interfaces are actually quite
power cost of storing information across tasks is relaxed, thus cost and power consumption are far
discussed in Chaps. 5–7. more important aspects than pure performance.
Both challenges are well addressed by tailoring
such circuits around the specific application. The
power consumption of the ADC is proportional to
1.3.3 On-Board Capabilities of IoT the datarate in Fig. 1.7 and the energy per conver-
Nodes sion in Fig. 1.6, and is plotted in Fig. 1.8. From this
figure, the power consumption of ADCs for IoT
IoT nodes need to have sensing, computation, nodes spans a very wide range, mostly because of
and wireless communication capabilities. In IoT the wide energy per conversion range in Fig. 1.6,
nodes design for a specific purpose, sensing can as dictated by the exponential relationship between
be typically made more inexpensive by tailoring resolution and energy (Freyman et al. 2014). This
the MEMS design and the analog interface confirms that tailoring the ADC to the specific
around the specific application. This permits to application is crucial in IoT, and the same consid-
substantially reduce the complexity that is expe- eration applies to most of the other building blocks
rienced by general-purpose platforms, and hence and sub-systems.
the cost. As simple example, Fig. 1.6 shows that Let us now consider the case where the raw
sensors for IoT applications cover a wide range sensor data is transferred directly to
of resolutions, hence using the appropriate ADC concentrators and cloud. Assuming a best-in-
resolution (see Chap. 13) is necessary to avoid class radio consuming 5 nJ/bit (ISSCC 2016),
using general-purpose platforms with over- the resulting power to wirelessly transmit such
1 IoT: Bird’s Eye View, Megatrends and Perspectives 11

Fig. 1.6 Resolution range required by various sensors

wireless power (W)

1E+8 5E-1
datarate (bps)

1E+6 5E-3

1E+4 5E-5

1 mW
1E+2 5E-7

1E+0
sensor
heart rate
humidity (capacitive)
battery monitor
temperature
accelerometer
magnetometer
altimeter/pressure
imager (VGA, RGB)
imager (MP4 compressed)
infrared proximity
gyroscope
microphone
CO2
light
strain
ultra-violet

Fig. 1.7 Datarate range required by various sensors, and wireless power required to continuously transmit data
(energy/bit assumed to be 5 nJ/bit (ISSCC 2016))
12 M. Alioto

ADC power consumption (W)

1E-03
1E-04
1E-05
1E-06
1E-07
1E-08
1E-09
1E-10
sensor
heart rate
humidity (capacitive)
battery monitor
temperature
accelerometer
magnetometer
altimeter/pressure
imager (VGA, RGB)
imager (MP4 compressed)
infrared proximity
gyroscope
microphone
CO2
light
strain
ultra-violet
Fig. 1.8 ADC power consumption under sampling rate associated with the datarate in Fig. 1.7, assuming an energy/
conversion step of 30 fJ

raw data is reported on the secondary y axis in communication. For example, the IoT node can
Fig. 1.7. From this figure, most of sensors cer- be proactive and monitor for critical or important
tainly exceed 1 μW and hence the range of prac- events (e.g., the crossing of a threshold, or an
tical IoT node power targets mentioned above. increase rate larger than a pre-set value), and
Hence, mere computation offloading to transmit data only upon their occurrence. From
concentrators and cloud through raw data trans- Fig. 1.7, this is particularly crucial in applications
mission is not an option for IoT nodes operating involving large datarates, such as continuous
continuously. vision and audio sensing. In such applications,
For some environmental sensors (e.g., temper- more intelligence needs to be embedded in the
ature, CO2, light, UV), duty cycling discussed in IoT node, such as the ability to perform pattern
the previous subsection is applicable since recognition and classification. Other options to
measurements do not need to be taken continu- trade off computation and communication are in
ously, as the related phenomena exhibit slower the choice of the data representation and sampling
time constants. For such sensors, a duty cycle of approach (e.g., compressive sensing, including
percentage points reduces the average datarate computation in the compressive sensing domain
down to hundreds of tens of bits/second, and the (Shoaib et al. 2015)), as well as signal dimension-
power down to tens of nWs. Often times, the other ality reduction (e.g., in-node feature extraction,
sensors cannot be duty cycled as the dynamics of which is equivalent to compression, with the fur-
the related phenomenon does not really allow it ther advantage that it is often a necessary task to
(e.g., accelerometers, gyroscopes, imaging, be performed anyway in many algorithms).
audio). In these cases, further power reduction From the above considerations, the wireless
can be achieved by leveraging the well-known power is always an issue in IoT nodes, and hence
computation-communication tradeoff (Min et al. requires the choice of appropriate communica-
2001), moving computation onto the IoT node tion standard for the intended range and datarate,
to reduce the volume of wireless data as will be discussed in the next section.
1 IoT: Bird’s Eye View, Megatrends and Perspectives 13

1.3.4 User Constraints The above challenges need to be addressed

through platform-based design and moderate
Other important requirements of IoT nodes come reconfigure-ability to reduce the design cost and
from the user, and are mainly related to cost and widen the range of targetable applications, espe-
security. Regarding the cost, consumer cially in consumer electronics. On the other
applications dictate a target of approximately 1 hand, such challenges are mitigated in
$/node as was discussed in Sect. 1.1 (which limits applications where the ability of directly receiv-
the die cost to a fraction of it). This clearly puts ing information on large numbers of objects is
pressure on the financials of the semiconductor particularly valuable. For example, this is the
industry due to the limited room for profit margin, case of manufacturing, logistics and smart cities,
and can be addressed through large sales volumes whereas it is not for single users in a smart home.
(say, at least several tens of millions per year) and Security is another important requirement
specialized hardware for reasons related to cost, coming from the user, as the IoT offers a very
power and form factor (see previous subsections). large number of backdoors to attackers, in view
On the other hand, achieving such volumes is diffi- of its large scale. In addition, traditional solutions
cult even deploying an IoT with a trillion devices. to counteract cyber-attacks (e.g., firewall, cryp-
Indeed, the IoT space is highly fragmented, and tography) are not applicable to IoT nodes, due to
only few applications are so pervasive that they their very limited power budget and cost. As fur-
require such large volumes (https://www. ther challenge, the dispersed deployment of IoT
mckinsey.de/files/mckinsey-gsa-internt-of-things- nodes makes it hard to keep track of individual
exec-summary.pdf). Similarly, it is hard to justify nodes, thus exposing them to physical attacks.
the non-recurring engineering cost of a new chip Such challenges require novel security
design for applications that do not require more approaches that embrace the hardware level rather
than tens of million pieces. This will require the than being confined at the network or software
development of an ecosystem that favors design level, in order to reduce the cost and energy, and
reuse, and platform-based design approaches. assure chip-level authentication (see Chap. 8).
Regarding the recurring costs, IoT node cost
reduction certainly requires more aggressive
on-die integration to limit the cost of off-chip 1.4 Looking at the Past: IoT
component assembly and testing. For example, as Natural Outcome
circuits for power delivery and harvesting need to of Technological Trends
avoid off-chip passive components (see Chaps. 10
and 11), and innovative integration techniques and The IoT can be shown to be a natural conse-
packaging becomes crucial to assemble heteroge- quence of historical trends that are relevant to
neous components in an inexpensive and ultra- its distinctive features, such as size, energy, sales
compact manner (Heterogeneous Integration volume, cost, with other software implications
Roadmap) (see Chap. 16). discussed at the end of the section. Other
As an additional challenge, IoT nodes are considerations on the evolution of the communi-
required to have a long lifetime (e.g., decades), cation infrastructure will be made in Sect. 1.7.
which translates into a missed opportunity to The Bell’s law observes that a new computer
replace the nodes for a very long time. In the class has appeared every 10 years, thus bringing
long run, this will expectedly make the IoT exponential improvements in computer size
market very different from the consumer mar- (100 smaller every 10 years) and cost (Bell
ket, which typically relies on periodic new et al. 1972; Bell 2008; Fojtik et al. 2013), as
waves of demand stimulated by incoming summarized in Fig. 1.9. This has driven the com-
generations of products with improved features puter market expansion in computer units by a
(and predictable release timeline, which allows factor of 10–20 every 10 years (Tsai 2014), as
planning). in Fig. 1.9. Based on the current dominant wave
Random documents with unrelated
content Scribd suggests to you:
concrete work; except that the drain pipe used will be paid for per
linear foot at the price named in the contract which price shall cover
the cost of furnishing and putting the pipe in place complete.
 PART II
INSTRUCTIONS TO INSPECTORS ON
STREET PAVING WORK
 INTRODUCTORY NOTE

Any general code of instructions for inspectors on street paving

work must necessarily be little more than tentative or suggestive, as
applied to any one city.
That here presented is intended for use in connection with the
specifications in Part I of this book. Different specifications might
make necessary many changes in the instructions that follow. The
local or special conditions in any city might also make them
inapplicable without material modifications and additions.
The author believes that in every city where considerable street
pavement work is prosecuted and a number of inspectors employed
on such work, some such code of instructions to inspectors should be
formulated and used. Even where the inspectors are of high
character and fully competent, the adoption of a system of rules and
regulations relating to their work will tend to unify procedure,
prevent misunderstandings and promote good discipline.
The preparation of such a code of instructions requires no little
time and thought, especially where there are no precedents to
suggest what is needed, or to be used as rough patterns. It is with the
hope that the code here presented may be found useful in this way,
even though it may be not appropriate for adoption in any given city,
that it is offered to municipal engineers.
 GENERAL REMARKS ABOUT INSPECTION
AND INSPECTORS

The necessity of having competent and honest inspectors upon any

public work during its construction need not be here argued. It is
generally understood and recognized in the case of work carried out
by contract. Where the contractor is incompetent, careless or
unreliable, inspection is absolutely necessary if good work is to be
secured. Even where the contractor thoroughly understands his
business and intends to faithfully comply with the specifications and
to do all his work in a proper manner, the inspector cannot safely be
dispensed with. The contractor cannot at all times be on the work;
whatever may be his intention and instructions, foremen and
laborers cannot be depended upon to exercise the proper care and
judgment, even where they have no inclination, motive or interest to
do otherwise, which is not always the case. It seems to be deeply and
almost irradicably fixed in the mind of the average foreman on
municipal work that he is expected to show his ability and skill by
evading or circumventing the strict requirements of the
specifications and the watchfulness of the municipal agents,
whatever his employer may say to the contrary. Even when fairly
careful, honest and conscientious he may need friendly oversight. He
is naturally and properly anxious to save all the money he can for his
employer, whether from a desire to loyally serve that employer’s
interest, or to enhance his reputation for doing work cheaply. He
may not appreciate the importance of minor requirements of the
specifications and may believe they can be ignored without real
detriment to the utility of the completed work. Like other men he
may sometimes be careless or forgetful, however good his intentions.
These human qualities are not confined to foremen on contract
work. They are liable if not likely to be found in the foremen on work
done for the municipality by the direct, or day’s work system. They
may be encouraged by the impression or belief that they will not be
held by the municipal authorities to as strict an observance or
accountability as they would be if the work were being done by a
contractor—which is too often true. The absence of some of the
motives of foremen employed by contractors is, therefore, not a
sufficient reason for dispensing with inspectors on work done by the
city direct.
It should be trite to say that inspectors should be chosen with
some regard to their qualifications for the work they are expected to
do. One cannot, of course, expect to secure experts for such positions
at the rate of compensation usually paid; but it is reasonable to
require that an appointee to such positions shall possess the
following qualifications:
That he shall be honest, trustworthy and loyal.
That he shall be a man of at least average intelligence and common
sense.
That he shall have some practical knowledge of or experience in
the work he is employed to inspect.
That he shall be a man of good personal habits.
That he shall be habitually punctual, industrious, and alert, and
shall possess a fair share of that quality commonly called “backbone.”
It will not, I am sure, be denied that a very large number of the
inspectors found upon municipal public work fall short in one or
more of these qualifications.
Inspectors who habitually expect or accept compensation or
gratuities from the contractor violate the first requirement, and their
number is astonishingly large. These are useless, or worse than
useless, to the municipality. There are not a few contractors who will
take advantage of the presence on the work of such inspectors to do
things that they would not do if thrown wholly on their own honor
and responsibility. Contractors can hardly be blamed for expecting
some compensation for the money thus exacted or accepted by the
inspector from them.
The requirement that the appointee shall have some practical
knowledge of and experience in the work he is expected to do, while
often disregarded, is of real importance. Good workmanship
embraces many small details that are essential, and inferable from
the specifications, but cannot be stated at length therein, and the
inspector should have a good working knowledge of these.
Unfortunately, large numbers of inspectors are employed for
personal or political reasons, or because more competent men are
not available. They know little or nothing from practical knowledge
and experience about the work they are assigned to inspect. Efficient
service cannot be expected from such, even if they possess the other
necessary qualifications. The obvious remedy, where a sufficient
number of qualified inspectors cannot be obtained at call, would be
to train up a permanent force by having the promising tyros serve a
sufficient apprenticeship under older and more experienced
inspectors; but this is seldom done.
Inspectors, whether appointed by the engineer in charge or not,
should be absolutely under his control, including the power of
dismissal without unnecessary formalities. Where they are under
civil service regulations and can be dismissed only upon charges and
after a hearing before some third person, the obstacles in the way of
promptly getting rid of an incompetent or unfaithful inspector are
often so great as to be discouraging if not deterrent. He may be
clearly below a reasonable standard of honesty or efficiency but it
may be very difficult to formulate distinct charges and support them
with positive evidence. Thus, one may be morally certain that an
inspector is accepting gratuities from a contractor, but unable to
prove it by legal evidence. If in such cases the judgment and integrity
of the responsible engineer cannot be trusted, he is not fit, himself,
for the position he occupies.
The question of how much authority and power shall be given to
the inspector is always a troublesome one. If he be too much limited
or restricted, he cannot properly discharge his duties or command
the necessary respect from the contractor and his employees; if he be
given too much authority and is disposed to be arbitrary or
unreasonable, the contractor may be unjustly treated. So far as the
definite requirements of the specifications are concerned there can
usually be no room for going astray, but in the numberless little
details not specifically covered in the specifications, or where their
language and intent is not clear, differences and disputes may result
which require a degree of knowledge and judgment that the
inspector may not possess, and where his decisions might be against
the interests of the municipality or unjust to the contractor. In such
cases the matter should always be referred to the engineer or his
assistant.
Where inspectors are employed upon work done directly by the
municipality the relation between them and the superintendent or
foreman should be clearly defined and understood. Briefly stated,
this relation should be the same as that between the inspector and
the contractor on contract work.
 INSTRUCTIONS FOR THE GUIDANCE OF
INSPECTORS ON STREET PAVING AND
INCIDENTAL WORK

GENERAL

1. The chief duty of the inspector is to see that the work to which
he is assigned shall be carried out in full and strict accordance with
the plans and specifications therefor, and with such additional
instructions as may from time to time be given by the engineer.
2. These instructions are intended to supplement but in no case to
take the place of the specifications for the work, which must be fully
and strictly complied with, unless they are changed or modified by
the engineer.
3. Wherever the words “the engineer” are used in these
instructions they refer to the chief engineer in charge of the work for
the municipality, or his authorized deputies or agents.
Wherever the words “the contractor” are used they refer to the
person, or firm, or corporation under contract to perform the work to
which you are assigned, or his agents or employees engaged upon the
work.
4. You will report to and be under the exclusive direction and
control of the engineer.
5. Upon being assigned to any work you will obtain copies of the
plans and specifications under which it is to be done, and will
carefully read and acquaint yourself with all their provisions and
requirements. If any part of these plans and specifications are not
clearly understood by you, you will ask the engineer for explanations
and instructions, and will be governed thereby. Failure to
understand or to be familiar with the plans and specifications will in
no case be accepted as an excuse for not complying with and
enforcing them.
You will also obtain, study and familiarize yourself with all the
blank forms, reports, etc., that are to be used by you on the work, so
that they may be correctly utilized; you will also obtain a proper
supply of such apparatus, tools, stationery, etc., as you will need
upon the work.
6. In inspecting any work done under contract you are to assume
that the contractor is bound to carry out in good faith the plans and
specifications attached to his contract, and that he may be, and must
be, held to a strict compliance with them unless you receive
instructions to the contrary from the engineer. You will be held
responsible for the enforcement of these plans and specifications so
far as they relate to the doing of the work.
7. In case the contractor or his agents shall neglect or refuse to
comply with the plans and specifications or to perform the work in a
proper manner, or in accordance with the instructions of the
engineer, you will direct him to suspend the particular work in fault
and will report all the facts promptly to the engineer and await his
instructions, which, when received, you will proceed to enforce.
8. You will be expected to be upon the work at all times when work
is in progress, whether during ordinary working or office hours, or
not. But in case the contractor shall work continuous day and night
forces, or otherwise so conduct the work as to require the presence of
an inspector for unusual or unreasonable hours of service, you will
notify the engineer and ask instructions or assistance. If for any
reason you are to be unable to be present whenever work is in
progress you will promptly notify the engineer in time to enable him
to supply a substitute.
9. All reports and records required of you by the engineer must be
promptly made up and forwarded to him at the time and in the
manner he may direct. You will promptly acknowledge in person or
writing any instructions received from the engineer and in case such
instructions are not clearly understood by you, you will ask for
explanation or advice.
You will keep a notebook or diary in which all occurrences or
matters of interest relating to the work will be recorded each day,
which book will be delivered to the engineer on the completion of the
work, or whenever your connection therewith may end. All diaries,
notebooks and records kept on the work are the property of the city.
10. While your first duty is to see that the work is done in
accordance with the plans and specifications, you will, so far as
consistent with this duty, be considerate of the interests of the
contractor and facilitate his work as much as practicable. Materials
should be inspected promptly upon delivery and in such a manner as
not to delay or discommode the contractor more than is necessary.
Errors, defects or delinquencies in the work should be detected and
called attention to promptly, so as to avoid tearing up and re-
construction as far as practicable.
11. Orders and instructions should be given to the contractor or his
foremen or representatives on the work and not to the laborers,
unless in urgent cases.
12. You shall not at any time, before, during, or after the
completion of the work, ask or accept from the contractor or any of
his agents or employees, or from any other person whom you may
have reason to believe is acting for the contractor, any money, gifts,
gratuities or other considerations of value, whether as compensation
for services rendered him or otherwise. The penalty for violation of
this rule will be dismissal from the service followed by such legal
prosecution as the engineer may direct. This rule does not apply to
the ordinary small courtesies common between gentlemen associated
in business.
13. It is a condition of your employment that your services shall be
loyal to the interests of the municipality, efficient, and satisfactory to
the engineer, and that whenever in his opinion your service falls
short of this standard he may dismiss you at any time, without
previous notice.
 STREET PAVING WORK

14. Sub-foundation.—The usual procedure in preparing the

sub-foundation for a pavement is:
The removal of the old pavement if one exists on the street.
The setting of new or the resetting of the old curbing. (If concrete
combined curb and gutter is used it is not constructed until after the
rough grading is done.)
The rough grading of the street.
The fine grading of the street.
In this work the main things for the inspector to look after are the
following:
If an old pavement is to be taken up and any part of it is to be
reserved for use by the city, see that the material is not injured in
handling and is disposed of as the specifications or the instructions
of the engineer require.
15. Curbing.—New curb stone should be inspected as soon as
practicable after it is delivered on the street so that the contractor
may have time to replace any that may be rejected. Rejected stones
should be plainly marked on their tops with red paint, and the
contractor required to remove them from the street. Stones that are
defective in dressing should be given some distinctive mark and the
attention of the contractor called to them so that the necessary
redressing may be done well in advance of the setting, and
reinspected.
If the old curb stone is to be redressed and reset, it should be
inspected as soon as possible after it is taken out and all stones that
are not suitable for re-use rejected and removed from the work. The
redressing of the stones should be looked after carefully. The
tendency in this work is toward the use of old stones that do not, or
cannot be made to comply with the specifications, and toward laxity
in the dressing.
Setting Curbing.—In the setting of curbing the points to be
looked after are:
To see that the trenches are excavated to the full dimensions so as
to provide space for the use of the full quantity required of concrete,
gravel or broken stone in which the curb is to be embedded.
To see that the stones are set to true grade and alignment.
That the concrete, gravel or crushed stone used is of the proper
quality, the proper quantity used, and that it is placed and
compacted so as to fill the spaces designated under and around the
stone. The tendency of the workmen is to treat these details as
unimportant and to slight the work.
After the curbing is set any spaces between it and the edge of the
sidewalk are to be cleaned out, filled and trimmed with Portland
cement mortar.
16. Rough Grading.—If plows and scrapers are used in the
rough grading, see that they are not allowed to penetrate or to
disturb the natural soil below the true grade, and to prevent this it is
better to require that the plowing and scraping shall not approach
nearer than one or two inches to the true grade.
If embankment or filling is necessary, the filling material must be
deposited in regular layers not more than six inches deep and each
layer must be well compacted by rolling or ramming before the next
layer is applied.
17. Fine Grading.—The fine grading (with pick or mattock and
shovel), to bring the sub-grade to the true grade and contour
required, should be done with reasonable accuracy and care.
Allowance must be made for compression under the roller, so that
when completed the surface shall not vary more than half an inch
below nor more than a quarter of an inch above the true grade, in
order that room for the full thickness of concrete foundation may be
provided, and no more. Humps, depressions and irregularities of the
surface are to be avoided. The fine grading will be completed by
thoroughly rolling or tamping the surface to make the soil firm and
dense. On some soft or wet, clayey soils, excessive rolling may
produce a plastic or “liver-like” condition and the rolling should be
discontinued where such a tendency appears.
18. Work Incidental to Grading.—Where it is a part of the
contractor’s work to reconstruct or readjust catch basins, manholes,
covers, and other structures, he should be required to keep this work
done well in advance of the laying of the concrete foundation. If this
work is delayed until after the main body of the concrete is in place,
leaving openings to be filled later, hurriedly and perhaps carelessly,
it may result in weak patches in the concrete around these structures,
where, in fact, the concrete should be strongest.

CONCRETE FOUNDATIONS

19. The concrete foundation of a pavement is a most important

part of the work, and must not be slighted in any way.
20. Quality of Materials.—The quality of the materials used for
making the concrete are fully described in the specifications, and
they must be substantially and uniformly adhered to.
The Portland cement delivered on the street will presumably have
been previously inspected and accepted by the engineer, but any
change in the kind, or brand, or apparent quality of the cement
delivered after the beginning of the work should not be permitted
without the approval of the engineer. Barrels or bags of cement not
plainly marked with the brand, or name of the manufacturer must be
rejected. You will frequently make balls of stiff mortar from the
cement and sand used (1 part cement, 2 parts sand) and set them
aside for observation. If in eight hours these balls do not set up hard
the fact should be promptly reported to the engineer.
The sand and stone will naturally vary slightly from time to time,
but should not differ materially from the requirements of the
specifications or the standards adopted by the engineer. Stone or
gravel that contains any considerable quantity, say more than 5%, of
foreign matter, such as soil or refuse, should be rejected. Stone
containing an excessive quantity of fine fragments or screenings
bunched together in the pile, should be mixed with the mass of larger
stone before it is used, so as to produce a mass of fairly even
composition. Stone that is coated with mud, or that is not sound and
hard should be thrown out.
Sand that is moderately coarse and does not contain more than
7½% of clay or soil, and is free from vegetable refuse, may be used.
21. Storing Materials.—Cement stored on the street must be
stacked up on dry lumber floors at least four inches from the ground
and effectually covered to protect it from rain or snow. Packages of
cement which, when turned out for use contain hard lumps should
be rejected.
Sand and stone must be stored on tight lumber floors to prevent
their admixture, when being shoveled up, with street soil or other
impurities. This requirement must be enforced.
22. Proportioning Concrete.—The proportions of cement,
sand and stone or gravel named in the specifications must be strictly
adhered to.
A barrel of Portland cement will be considered as four cubic feet
and a standard bag of cement as one cubic foot.
It is usual for the contractor to measure the sand and stone by
wheelbarrow loads. In that case you will measure and determine the
capacity, in cubic feet, of the convenient or average wheelbarrow
load, and determine the size and number of loads each of sand and
stone required for one batch of concrete (if made by hand) and
thereafter see that the quantities thus determined are used, and the
wheelbarrow loading uniform. To measure the contents of a
wheelbarrow load, use a square box holding one cubic foot. It cannot
be accurately measured on the barrow.
As a check upon the proportioning of materials, you should
occasionally keep a record of the quantities of cement, sand and
stone used for, say, one or two hours, and mark and measure up the
quantity of concrete made with them, and from this data compute
the ratios of the materials used. (In ordinary street concrete,
proportioned 1:3:6, the volume of rammed concrete made will just
about equal the volume of the stone used.)
Where the concrete is mixed by a machine with automatic feeding
devices, reliance must not be placed upon the machine for proper
proportioning. Frequent tests of the materials used and the volume
of concrete turned out should be made as described above, and any
necessary corrections made in the feeding devices of the machine.
This test should be made daily if practicable. The eye will usually
detect any material change in the composition of the concrete, and
whenever such changes are observed the composition ought to be at
once checked up by measurement.
23. Mixing Concrete.—See that the specifications are followed.
Use sufficient water to make what is commonly called a wet concrete
but not so wet that free water will drain from the mixed batch before
ramming. The test for proper mixing will be that all the fragments of
stone are completely covered with mortar.
24. Placing Concrete on Street.—All concrete must be placed
and rammed before it begins to stiffen or set. Concrete should be so
handled that the mortar will segregate or separate from the stone as
little as possible. When in place on the street it should make a mass
of uniform consistency and appearance. The concrete should, as soon
as placed on the street, be graded to a proper thickness and surface,
and then well compacted by ramming. Insist on good ramming; this
part of the work is not often properly done. The surface of the
concrete should be brought to the true grade by proper grading and
tamping. Where necessary to raise the surface, additional concrete,
made of small stone, may be applied during the tamping. Neat
mortar should not be used for this purpose. Sweeping mortar over
the surface with street brooms should not be permitted—it is useful
only to cover up defects and to produce an extra smooth surface
(which is not usually desirable). To measure the thickness of the bed
of concrete, lay a straight-edge upon the surface with the end of the
straight-edge projecting a foot or two beyond the edge of the
concrete, and measure the space from the lower edge of the straight-
edge down to the sub-foundation.
Care must be taken to finish the surface of the completed concrete
at the proper distance below and truly parallel to the grade and
contour of the finished pavement. This is more important in the case
of sheet asphalt pavements, since irregularities in the concrete will
cause variation in the thickness of the finished asphalt surface, which
will affect the durability of the pavement. The grade stakes for the
concrete should therefore be carefully set and worked to, and humps
or depressions between the stakes should be avoided.
 25. In hot, dry, weather the concrete must be kept damp by
watering it until it is at least five days old, and horse-travel over it
must be prohibited until it has set up hard enough not to be injured
thereby.
26. Old Paving Stone Foundations.—If a foundation is to be
made of old paving stone blocks reset, the important points to be
looked after are,
The preparation of the sub-foundation.
The sand bed.
The setting of the blocks.
The grouting of the blocks.
The sub-foundation should be prepared, where necessary, as
stipulated in Sects. 16 and 17 of these instructions. This part of the
work is likely to be carelessly done unless proper work is insisted
upon.
The sand used as a cushion course must be of good quality and
reasonably clean. The use of mixed sand and soil taken from the
street excavation should not be permitted.
The blocks, after being cleaned from adhering dirt, should be set
with reasonable care, well bedded in the sand cushion, and laid at
such an elevation below the completed pavement surface, that after
being well rammed, the surface will be at the proper elevation below
the pavement datum.
The grouting of the joints is the most important part of the work.
See that the mortar is made and used in accordance with the
specifications, and that the joints are completely filled.
27. Broken Stone Foundations.—The specifications for
broken stone foundations are sufficiently explicit, and compliance
should be insisted upon. Thorough rolling is especially important.

SHEET ASPHALT PAVEMENTS

28. The manufacture of the mixture for sheet asphalt pavement,

and the laying of it on the street, involve a number of separate
elements or operations. The quality and durability of the pavement
will depend largely upon the skill, intelligence and accuracy with
which each and all parts of the work are carried out. The inspector,
both at the mixing plant and on the street, must give constant and
intelligent attention to the work.
29. Inspecting at Paving Plant.—Inspectors at the plant where
the asphalt paving mixture is made will be provided with the
following apparatus:

1 nest of standard sieves, Nos. 10, 30, 50, 80, 100 and 200
2 thermometers for sand
1 tank thermometer
2 standard flow-plates and one dozen cartridges for same
1 balance for weighing sand
6 dishes, glass or porcelain, 4 inches in diameter.
30. The more important things for the Plant Inspector to watch
during the manufacture of asphalt paving mixture are:

a. Quality and grading of the sand.

b. Temperature of heated sand.
c. Uniform quality of the refined asphalt and flux.
d. Temperature of melted bitumens.
e. Uniform consistency of asphaltic cement.
f. Correct ratios of sand and asphaltic cement.
g. Thorough mixing of materials.
31. Sand.—The general quality and grain-size of the sand,
required by the specifications, must be closely watched and adhered
to.
It is hardly ever possible to secure a natural sand that will conform
strictly to the theoretical composition, but substantial compliance
should be required unless the Chief Engineer directs otherwise. After
any sand has been approved for a given contract or street, all the
sand thereafter used for that street should conform closely to the
standard so approved; for any material change in the quality of the
sand will change the quality of the pavement produced, unless the
proportions of the other materials are changed accordingly. Any
material change in the sand should therefore be promptly reported to
the Chief Engineer.
32. Sand for the sieve tests should be taken from the hot-sand box
and allowed to become cold. From the sample thus taken, weigh out
50 grams. Arrange the nest of sieves with the No. 200 at the bottom
followed by Nos. 100, No. 80, No. 50, No. 30 and No. 10 at the top.
Place the weighed sample upon the No. 10 sieve and thoroughly
shake the nest until all the sand that will, has passed through the
various sieves; then weigh the various portions, beginning with that
which has passed the No. 200 (deducting in each case the weight of
the dish). The portion remaining on the No. 200, is the quantity
passing the No. 100, etc. The sum of all these portions should be 50
grams, though it will usually be slightly less because of unavoidable
waste. In use, keep the sieves clean and as free as possible from
adhering grains by using a bristle brush.
33. Temperature of the Sand.—The temperature of the heated
sand should be frequently noted by plunging the thermometer into
the mass. It should not fall below 315° F. nor exceed 375° F. It is very
important to avoid over-hot sand, because, in the mixer, the very
thin films of bitumen covering the hot-sand grains are readily and
quickly scorched by the excessive heat in the sand.
34. Refined Asphalt and Fluxing Oil.—Unless otherwise
notified by the Chief Engineer you will assume that the stock of
refined asphalt on hand at the beginning of the work has been
inspected and is satisfactory. You will, however, note the appearance
of the material and the character and marking of the barrels or
packages, and will not permit the use of any new shipments of
materials of different appearance or brand without authority from
the Chief Engineer. The above instructions apply also to the oil used
for fluxing.
35. Ratio of Asphalt and Flux.—You will frequently note and
record the weight of asphalt and of flux used in charging the melting
tanks.
36. Temperature in Melting Tanks.—The melting kettles are
usually equipped with fixed thermometers, but their reading should
be frequently checked with your detached tank thermometer. You
will see that the temperatures in the kettles are kept within the limits
named in the specifications. See that the melted asphaltic cement is
kept constantly agitated as required by the specifications.
37. Uniformity of Asphaltic Cement.—It is very important
that the asphaltic cement shall be of uniform composition and
consistency, or hardness.
When the composition and consistency of the asphaltic cement to
be used for the surface-course of any street has been determined
upon and approved by the engineer, you will take out and preserve a
sufficiently large quantity to be used thereafter as a standard, and
each subsequent tank of cement, before it is used, should be tested
by comparing it with this standard in the following manner, using
the flow-plate apparatus:
Melt and fill two cartridges with the standard cement, and two
with the cement to be tested, marking them for identification. When
cold, carefully remove with a wetted dull knife any surplus cement
from the ends of the cartridges so as to leave them exactly full and
the metal ends clean. Immerse them in cold water until they are all of
the same temperature. Then place and fasten these filled cartridges
on the grooved plate with their lower ends exactly to the zero line on
the plate, the standard samples alternating with the new. Then place
the plate, at an inclination of about 45°, in front of a source of
uniformly distributed heat, as, for instance, near a steam boiler or
large steam pipe, or (when the weather is favorable) in the sunshine,
facing the sun. Leave the plate in this position until the bitumen in
the cartridges shall have melted and flowed down the grooves an
average of from four to five inches. Then measure the lengths of flow
of each from the zero line on the plate. If the average flow of the new
samples differs more than 5 per cent. from that of the standards, the
new cement should be regarded as unsatisfactory and should be
corrected by adding more asphalt or flux, as may be required.
(If a penetration apparatus is supplied or is available, the
penetration method may be used instead of this flow-method for
determining the consistency of the cements. Proper instructions for
its use will accompany the apparatus.)
38. Ratio of Sand and Cement.—When the proper ratio of
sand and asphaltic cement has been determined for any street you
will see that this ratio is carefully adhered to. The correctness of the
weighing scales used should be frequently tested.
39. Mixing.—When the sand and asphaltic cement are properly
mixed the appearance of the mass will be perfectly uniform
throughout, with every grain of sand entirely and evenly coated with
cement. The mixing should, however, be continued somewhat
beyond the attainment of this condition, so as to insure
thoroughness. After the proper result has been arrived at, either the
time in the mixer, or the number of revolutions of the mixer shaft
should be noted and the standard thus determined substantially
adhered to thereafter. With the standard mixers commonly used, the
hot sand and the filler (dust) should first be put into the mixer and
mixed for about fifteen seconds; then the asphaltic cement added
and the mixing continued as long as necessary (in the usual mixer
not less than sixty seconds).
40. Records.—You will keep complete records of the results of
your observations and tests in a book which will be provided for the
purpose, and will make such reports to the Engineer as he may
direct.
41. Inspection on the Street.—The leading things to be kept in
mind by the street inspector, in the laying of asphalt pavement on
the street, are the following:

a. Temperature of mixtures when applied to the street.

b. Proper preparation of the street surface before material is laid.
c. Laying and rolling of the binder-course.
d. Laying of the surface-course.
e. Rolling the surface-course.
42. Inspector’s Equipment.—The inspector will be supplied
with the following apparatus:

1 armored thermometer reading from 200° F. to 500° F.

1 graduated spatula.
1 twelve-foot straight-edge.
1 fifty-foot tape line.