This edition first published 2020
© 2020 John Wiley & Sons Ltd
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by law. Advice on how to obtain permission to reuse material from this title is available at http://www.wiley.com/go/permissions.
The right of John Davies and Carolina Fortuna to be identified as the authors of the editorial material in this work has been asserted in accordance with law.
Registered Offices
John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA
John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK
Editorial Office
The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK
For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com.
Wiley also publishes its books in a variety of electronic formats and by print‐on‐demand. Some content that appears in standard print versions of this book may not be available in other formats.
Limit of Liability/Disclaimer of Warranty
While the publisher and authors have used their best efforts in preparing this work, they make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives, written sales materials or promotional statements for this work. The fact that an organization, website, or product is referred to in this work as a citation and/or potential source of further information does not mean that the publisher and authors endorse the information or services the organization, website, or product may provide or recommendations it may make. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for your situation. You should consult with a specialist where appropriate. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.
Library of Congress Cataloging‐in‐Publication Data Applied for
HB ISBN: 9781119545262
Cover Design: Wiley
Cover image: Bangkok cityscape © Mongkol Chuewong/Getty Images, Internet of things © MONOPOLY919/Shutterstock, Business data market elements © Macrovector/Shutterstock
Professor John Davies is Chief Researcher in BT's Research & Innovation Department, where he leads a team focused on Internet of Things technologies. He has a strong track record of researching and innovating and his current research interests include the application of Internet of Things and semantic technologies to smart cities, smart transport, business intelligence, and information integration. He currently leads BT's contribution to the UK flagship Manchester‐based CityVerve IoT smart city programme as well as the H2020 NRG‐5 Smart Energy project and he co‐wrote the Hypercat IoT standard. John has authored several technical books and written over 90 scientific publications. He is the inventor of several patents. He is a Fellow of the British Computer Society and a Chartered Engineer. John is a visiting professor at a number of UK universities and holds a PhD in Artificial Intelligence from the University of Essex, UK.
Carolina Fortuna is a Senior Research Fellow at the Jožef Stefan Institute. She received her BSc in 2006 from the Technical University of Cluj‐Napoca, Romania, her PhD in 2013 in Slovenia, was a postdoctoral research associate at Ghent University 2014–2015 in Belgium, and a visiting researcher at Stanford University, USA, in 2017. Her research is interdisciplinary focusing on semantic technologies with applications in modeling of communication and sensor systems and on combining semantic technologies, statistical learning and networks for analyzing large datasets. She has participated in multiple collaborative research projects, taking leadership roles in several. She has co‐authored over 50 peer‐reviewed publications and gained industry insight by working with Bloomberg LP (New York) and Siemens PSE (Romania).
The Internet of Things (IoT) is the next phase in the evolution of the Internet and will transform our business and personal lives in many areas. IoT refers to the increasing trend for many types of objects including vehicles, environmental sensors, traffic sensors, clothing, and all kinds of consumer goods to be connected to the Internet and to have the ability to sense, communicate, network, and produce new information. The information generated by IoT devices is already exploited in a wide range of early applications: there are existing uses in the field of transport, smart cities, retail, logistics, home automation, and industrial control among others.
The IoT will generate massive volumes of data that flow to the computers for analysis, resulting in the collection of much richer information and insights in real time and used by automated systems to respond intelligently with appropriate actions.
The IoT has been a recurrent theme among commentators since the term was coined in the late 1990s. It involves a radical new view of networked ICT and the relationship between information systems and the physical world. The vision of IoT is that any physical object can be given the ability to measure and respond to its environment and to communicate with other objects or with computer systems anywhere in the world. This is now becoming technologically feasible and commercially viable.
Recently, a number of technological and socioeconomic drivers for the IoT have emerged, leading to expectations of very rapid growth over the next 5–10 years.
A key factor is the falling cost of essential components that will turn things into connected devices. Gartner claims components such as Wi‐Fi radios, GPS chips, or microcontrollers will fall toward less than €1 each in the near term when purchased in volume.
Another driver is the fact that the efficient use of natural resources is becoming increasingly important – with a new emphasis on costs and security of supply, as well as concerns about sustainability. Governments and organizations are motivated to improve the efficiency of their operations and are looking to technological solutions. Urbanization continues to increase globally, and cities are an efficient way to structure societies where natural resources are constrained. The IoT has a central role to play in making more efficient and effective use of finite resources. For example, traffic congestion is now estimated to cost the UK economy more than £300 billion between 2013 and 2030. IoT applications providing intelligent management and improved communication with road users have the potential to reduce congestion – and hence pollution – significantly.
From a technological point of view, achieving the scale and connectivity envisaged for the IoT depends on the ability to deploy and operate connected sensors, controllers, and actuators cheaply and easily. Network access has rapidly improved, with a range of wired and wireless access technologies now providing a wide coverage. In addition to cellular and Wi‐Fi networks, low power radio technologies well‐suited to machine‐to‐machine communication patterns are becoming available. In addition, 5G networks with a much enhanced bandwidth and latency characteristics will be deployed in the near future. Flexible computing infrastructures such as Cloud are now well established and offer the ability to deploy applications on demand in any region of the world.
Worldwide, there are today approximately 1.5 billion PCs and more than 1 billion mobile phones connected to the Internet. The IoT will greatly expand the number of connected devices by facilitating a wide range of uniquely identifiable new devices to be connected. An often‐quoted estimate is that there will be 25 billion objects connected by 2020. These new networked devices can use the Internet to publish data about their status and to receive data from other devices and human users.
In short, the time is right for IoT.
IoT applications have a characteristic structure with three major layers: (i) collecting data from or delivering messages to devices, (ii) enrichment of data to generate information in context, and (iii) applications that process information and initiate appropriate actions. These are primary activities in the IoT value chain with common requirements (e.g. secure distribution, digital storage, and computational processing) that can be generically supported. This volume covers all the major technologies required to support this value chain, including connectivity, security, data privacy and trust, and a range of AI and related techniques for extracting valuable and actionable insight from the IoT data.
By reading the book, readers new to this area, albeit having some basic ICT background, will become well acquainted with the concepts, components, technologies, and application areas related to and enabled by IoT. Readers already having some acquaintance with IoT will deepen their insight into IoT technologies as well as novel application areas.
March 2019
Prof. Tim Whitley
Managing Director, BT Labs
Adastral Park
Martlesham
UK
The editors wish to thank Dr. Andrew Reeves for proofreading a number of chapters, Paul Deans for valuable input on graphical design, and Maruša Mazej for reworking the figures and thereby improving the appearance and accessibility of the book.
Chapters 11 and 13 are partly based on work done on the NRG5 project, which received funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No. 762013.
John Davies1, and Carolina Fortuna2
1British Telecommunications plc, Ipswich, UK
2Jožef Stefan Institute, Ljubljana, Slovenia
The physical world is becoming ever more closely connected to information systems as sensors and actuators are incorporated into a wide variety of physical objects – from highways to pacemakers to cattle to running shoes to factories – and then connected to the Internet via a range of wired and wireless networks. This is the Internet of Things (IoT) and it is already generating massive volumes of data. The result is that much richer information can be collected (in real time) and used by automated systems to provide actionable insight and to respond to changing contexts with appropriate intelligent actions. IoT has rapidly moved from the conceptual phase to widespread use in real‐world applications in recent years.
The IoT will deliver significant innovation in many different areas, including future cities, transport, health and social care, manufacturing, and agriculture. Sensors can now be deployed at low cost to instrument the world to a far greater extent than has been possible before. There is increasing recognition of the potential value in opening up data resources so that they can be exploited more fully.
At the highest level, many of the IoT applications being considered appear similar – involving the collection of information from a range of sensors and other sources, interpreting this in a specific context, and then making better decisions that improve a behaviour or a process. For instance, smart watches or other types of wearable sensing devices are able to drive improvements in our behaviour toward a healthier daily routine. Merchandise tracking sensors can lead to a better understanding of supply chains and deliver optimization of costs and minimization of carbon footprints. IoT has a unique potential for automating and improving man‐made systems and behaviours by enabling unprecedented understanding and insight. For example, IoT data enabled a recent comprehensive global study across 111 countries on the impact of physical activity variation and the built environment on health [1].
The IoT has been a recurrent theme among commentators since the term was coined in the late 1990s. The concept has evolved from early work on Radio Frequency Identifier (RFID) technology which represented a hardware related break‐through that aimed to connect everyday objects to a network. This perhaps constituted the first wave of the IoT, which then developed beyond the initial hardware world innovation, and focused increasingly on developing new types of sensors and sensing materials, as well as on developing new communication technologies and protocols. As a result, a wide variety of new communication technologies emerged in the early years of the twenty‐first century which were able to support the ubiquitous deployment of a wide variety of sensors. We refer to this as the second wave of IoT. In the last decade, the focus of IoT has shifted to data collection, processing and security aspects and this period is termed the third wave of IoT. This book focuses primarily on this most recent wave and covers all key aspects including data management, processing, and analytics as well as security, privacy and trust as depicted in Figure 1.1. Real‐world examples are given that show the application of IoT technologies in a number of different sectors.
Figure 1.1 The IoT ecosystem.
A number of different actors typically participate in any deployment of IoT technology and we will refer to this set of stakeholders and the relationships between them as the IoT ecosystem. Such stakeholders may play one or more different roles. These include sensor providers, connectivity providers, information providers, application developers, analytics service providers, platform providers, and end users of information and applications.
Information providers in IoT ecosystems are often owners of sensor deployments. The primary purpose of their sensors may be for their own use but they may choose to make some of their data available to others, either on a commercial basis, to meet their obligations (particularly for public sector organizations), or for the general good. Various data processing platforms may also be information providers, even if they are not directly associated with “Things.” We refer to these as derived information providers; while not being the primary source of any information, they create value by combining data from multiple sources, transforming or applying various analytical techniques. These additive data sources could include: contextual (e.g. geographical, administrative) information; notifications of events such as traffic incidents and sporting fixtures; or, perhaps, rare events such as anomalies in a production process.
In efficient IoT ecosystems, information providers should be able to easily publish their services or data resources and advertise their availability via an easily accessible catalogue so that potential users can independently discover and assess their utility. This scenario is perhaps similar to the app stores that are commonplace today to make applications easily available. It is important to note that making data available should not imply relinquishing ownership rights; consequently, information providers also need the ability to define access controls, together with terms and conditions for use of the data they publish.
Platform providers have a key enabling role in the IoT ecosystem. They do not directly provide information or build dedicated services or applications but support stakeholders in other roles by providing a set of functionalities that all can use. This allows other participants in the ecosystem to focus on their own core activities and helps to accelerate innovation in the ecosystem. Platform providers may provide computing and storage infrastructure, as well as analytics services, which could include artificial intelligence (AI) capabilities such as summarization, enrichment, and reasoning.
Each platform provider will use specialist hardware and software tools and offer general‐purpose frameworks that an end user can exploit to define their own workflows. For instance, an edge or cloud provider offers on demand compute and storage resources that can be configured and modified on demand by users. Certain platform providers, typically application domain experts, offer a more complete service, including consultancy services, to support end users who may not have the necessary systems, data science, or analytics expertise.
Application developers produce applications that process the available data within a specific context to produce actionable insight for end users. Application developers should be able to discover what data and platform resources are available to them, what the key features and costs of each resource are, and assess which ones meet the needs of the applications that they want to build. This includes both the information content of the resources and practical considerations for the resources, such as dependability (accuracy, availability, etc.), conditions of use, or commercial considerations.
End users participate in the ecosystem by using the information and applications that are made available to them by other stakeholders. The end users can be private persons or institutional decision makers. As the ultimate beneficiaries of the functionality provided by the other stakeholders, it is important that their experience is positive and the ecosystem delivers real value for them. An IoT ecosystem will not be sustainable without the trust of its end users.
For individual end users, participation in the ecosystem is generally via an application. Often this application will make use of information that is generated through their use of the application, for example the user's location is often used as a data source by applications on mobile phones. The situation where the individual is an information provider needs to be addressed with care, particularly where personally identifiable or potentially sensitive information may be involved. Open engagement with end users that ensures they are properly informed and understand that they are included in the ecosystem is essential.
Along with IoT, artificial intelligence (AI) comprises an increasingly pervasive and important set of technologies. Recent years have seen significant advances in AI in a number of areas [2]. IoT and AI are inevitably interconnected, given the vast volumes of rich data generated by IoT and the ever‐increasing capability of AI systems to analyze, extract insight, and make decisions from that data. Thus, any discussion of the role and impact of AI would be incomplete without consideration of the link to IoT and in this volume a number of chapters are included that discuss the role of AI in IoT systems.
The vision of the IoT is that digital systems can be given the ability to sense, process, and extract useful information and actionable insight from the world and respond to the environment accordingly (typically via actuation). From an AI and robotics perspective, we can make an analogy with human sensing/actuating capabilities and the five human senses that receive inputs from the external environment. These stimuli are sent to the brain via the nervous system, and finally the brain processes the stimuli as depicted in Figure 1.2. The result is typically information generation and in some cases is also the initiation of action: the brain transmits commands to muscles, which then trigger motion or speech, or another appropriate response.
Figure 1.2 Human versus IoT: the sense, process, and actuate analogy.
By analogy, the “things” in the IoT are the sense organs, which detect the stimuli. Devices featuring microphones detect sound; ones featuring gas sensors are able to detect gases such as volatile organic compounds; ones featuring cameras are able to record images or videos; ones featuring accelerometers are able to record motion and vibrations; and so on. The sensed data can then be processed locally on the devices (“edge processing” in the terminology of IoT) or sent via wireless or wired technology to data platforms (processing and storage engines, such as the Information Exchange depicted in Figure 1.1). This model is similar to how stimuli from sensory organs are sent via the nerves to the brain. These processing and storage engines then process the received information and generate actionable insight or other types of knowledge. In more advanced applications, systems can also initiate an action such as adjusting a setting in a heating system, sending a tweet, or actuating hardware controlling an industrial process.
One key distinction between the IoT and the way in which humans process and react to sense data is that, while in the case of the humans the sensors and processors are co‐located, in the case of IoT the system is typically distributed as shown in Figure 1.2. This analogy has inspired researchers and enthusiasts for decades, but, in spite of some reports in the media, relevant sensorics, robotics, and AI technologies are still far from achieving human capabilities.
There are two important classes of sensor‐based IoT applications – those that aim to monitor and respond to time‐sensitive conditions and those that collect data over a longer period of time for analysis of a longitudinal dataset. In either case, much of the time and effort involved can be spent on activities that are generic. It is advantageous to build new applications in an environment where these generic problems have already been solved by others, with robust solutions available to all.
A key technical aspect of IoT is the need to work at a very large scale (many devices, large volumes of data, and with ever‐increased scope for automation). We are also seeing increasing potential to share information much more widely. These needs are being driven by ever‐decreasing component costs and device miniaturization. As explained above, an IoT ecosystem consists of a number of independent stakeholders, all sharing a common interest in particular kinds of information and obtaining benefit from participation in the ecosystem. This could be as a commercial provider of information or analytic services, as an application developer or as an end user, for example.
Use of shared services and facilities generally involves a compromise – typically giving up some level of direct control in return for reduced costs. In the case of today's global communications networks (including the Internet), the case for common services is very strong. Cloud computing and storage are also becoming widely accepted, although there are still many situations where private infrastructure is preferred. The potential for an IoT ecosystem to stimulate and enable innovation is clear, but all participants need to have confidence in the value proposition and be convinced that it meets their needs. If this is not the case then a sustainable ecosystem will not be possible. Areas of concern for participants will include security and trust, respect for personal and commercial rights, dependability, performance, the ability to comply with legal and regulatory obligations, and cost. Predictability, simplicity, and flexibility are additional important characteristics.
In this book, we address the entire vertical technology stack of IoT, with special emphasis on the data aggregation, processing, management, analysis, and exploitation aspects.
Importantly, we also discuss recent developments in distributed trust, security, and privacy options. Currently we are at a critical point in the development of IoT. While there is a clear need for sensor‐based, data‐driven decision making with the potential for significant commercial and societal benefits, there are also increasing data misuse concerns that malicious users could abuse, with such a system causing unintended actions and destabilizing normal operation.
This book is comprised of two main parts. Firstly, it brings together a description of the full technology stack of IoT with, as mentioned, a focus on the data‐driven aspects. These include data modeling, processing, and security. There is also discussion on the critical related aspects of connectivity, privacy, and trust. The second part of the book explains how this technology is being applied in practice and the benefits that it is delivering by providing a number of chapters describing specific applications across a number of industry sectors.
The first part of the book can be seen as comprising three subparts. The first subpart, formed of Chapters 2 and 3, introduces data collection (connectivity) and computational infrastructure. The second subpart, comprised of Chapters 4 to 8 discusses the various aspects of data processing. Finally, Chapters 9 to 11 discuss security, trust, and privacy challenges as related to IoT.
More specifically, Chapter 2 analyzes connectivity options for the IoT, with particular focus on dedicated low‐power wide‐area network and cellular technologies. Enabling low‐power communications is perhaps the most important challenge for IoT devices, which are often battery‐operated. Chapter 3 introduces emerging edge computing architectures and technologies. Topics include data computation close to the network edge as well as the challenge of the efficient management of large numbers of devices through their lifecycle.
Chapter 4 discusses IoT data platforms and the need for data interoperability, so that data coming from various IoT systems can be more easily integrated for developing informed decision‐making systems and thereby maximizing the value of IoT data. Chapter 5 focuses on architectures and emerging technologies that enable the processing of streaming data. IoT deployments in several application areas, especially for fault detection in critical systems, should produce real‐time insights for alerting and decision making, meaning that specialized data stream processing systems are often required. Chapter 6 describes the important role of computer vision in IoT, particularly in drone‐operated scenarios using relatively lightweight computation. Chapter 7 introduces structured knowledge representation and reasoning technologies for IoT. This chapter can be seen as showing the suitability of symbolic AI applied to IoT. Chapter 8 then overviews the role of humans in crowdsourcing IoT data collection as well as data annotation and labeling for AI algorithms.
Chapter 9 discusses security challenges in an IoT world and provides general guidelines for preventing undesired events. Chapter 10 considers distributed ledger technology, also known as blockchain, as a possible trust enabler in an IoT ecosystem. Chapter 11 reviews data privacy standards, regulations, and technologies that are relevant for particular types of IoT‐generated data such as in the healthcare domain.
The second part of the book comprises three chapters focused on the application of IoT technologies in selected application areas: healthcare, energy, and air quality and road transportation. Chapter 12 shows the important role of IoT data representation, interoperability, and privacy in integrated digital infrastructures for hospitals. Chapter 13 shows the application of IoT technology and the need for trust and for real‐time processing systems in emerging smart grid energy systems. Finally, Chapter 14 discusses the role of IoT in optimizing road transportation for improved air quality.
We conclude by discussing the future outlook for IoT and related technologies.