Sign in or register
for additional privileges

C2C Digital Magazine (Spring / Summer 2019)

Colleague 2 Colleague, Author
Cover, page 19 of 23

 

You appear to be using an older verion of Internet Explorer. For the best experience please upgrade your IE version or switch to a another web browser.

Book Review: Understanding the IoT as Interdependent Technologies

By Shalin Hai-Jew, Kansas State University 




IOT:  Platforms, Connectivity, Applications and Services 
By Abdulrahman Yarali 
New York: Nova Science Publishers 
2018   340 pp.  


“It’s an understatement to say that developing nations have many more preconditions when it comes to the Internet of Things.  Before it is physically deployed, the first Internet of things (sic) related challenge that must be addressed is data regulations and frameworks. Ensuring that users of a completely interconnected network can have privacy, security and trust the integrity of their data is perhaps the most important underlying issue.  With world politics increasing in complexity and turmoil, protecting people’s privacy and security is the only way to prevent governments from turning such powerful networks into weapons of mass surveillance and infringement. These security concerns should be addressed with standards and be consistent across all nations in order to be effective.”  

--  Dr. Abdulrahman Yarali, Preface, 2018, p. xi 


Internet of Things (IoT) technologies have been applied to “smart cities” to further city design (for future cities); to enable increased security and situational awareness; to promote more efficient public transportation and management of vehicular flows; to provide more efficient package delivery, and other impacts.  Digital government endeavors, informed by citizen feedback and data, have continued to improved.  In private industry, the IoTs has enabled advances in unmanned aerial vehicles (drones), in robotics, in driverless cars and other motor vehicles, and in fly-by-wire aircraft, among others. 

More personally, in our smart homes, we engage with smart speakers to acquire information, make purchases, set up services, and interconnect with others. We acquire weather information from a world pocked with weather sensors.  We wear smart devices to track our biometrics and report on our fitness to programs.  Our homes—indoor and out—are wired with security cameras and panic buttons.  There are environmental controls applied to the living spaces for comfortable temperatures and clean air.  Our smartphones enable smart accesses to information, services, and technologically-enabled capabilities. 

In one sense, the Internet of Things (IoT) has been around for decades, in systems of interconnected devices in the management of industrial systems, security, surveillance, building environment controls, and others.  As a “thing” that has come to the forefront of the popular imagination, “IoT” is newer and is hot.  Dr. Abdulrahman Yarali’s new book IOT:  Platforms, Connectivity, Applications and Services seems designed to meet the latter need of the popular mind albeit with an overlay of light technological thinking.  

Yarali (2018), a professor of Telecommunications Systems at Murray State University in Kentucky, defines IoT as follows:  “a networked interconnection of computers, devices, and sensors that are capable of processing their data without depending entirely on people for input” (p. 2) Heterogeneous “smart objects” may be loaded with sensors, actuators, cameras, and other equipment with a range of functionality.  To achieve effective IoT, he suggests that much work still needs to be done, to set up the laws, data policies, respective technologies, processes, and public expectations, to enable an IoT that is centered around human (and animal) well-being in smart cities, environmental health and sustainability, governance, warfare, human health, human living, human learning, entertainment, and other areas.  




Figure 1.  Smart City Graph (by Dr. Sam Musa, Drawn by Cybersecurity 101 on Wikimedia Commons)


Technological Layering

The IoT is enabled through inter-related technologies, what Yarali’s title refers to as “platforms, connectivity, applications and services.”  The popular thinking is that hyper-connectivity from meshes and networks of interacting devices and information may enable new capabilities not possible otherwise.  Between developed and developing nations, the IoT may manifest differently.  In the latter, there are more “preconditions” to deploying the IoT based on costs, infrastructures, laws, security concerns, privacy concerns, and others (Yarali, 2018, pp. x and xi).  Onboarding IoT will result in greater load processing on existing technological infrastructures, and it will require additional energy to run billions of new devices using battery and other power.  The respective technologies will require adaptations.  Standards need to be set, and variances need to be accommodated.  He writes:  “Ideally, future networks would be backward compatible with existing infrastructure, but it may not always be possible. The infrastructure must be power efficient, scalable, intelligent, standardized and heterogeneous” (p. xi) The industrial IoT (IIoT) has also advanced in “the past two decades” and “will probably require another two decades to achieve its optimum potential” (p. 157).  To advance the field for the various potential applications of IoT, proper forethought in policy and investments and engineering need to be applied at every phase.  The 16 chapters of this text offers some light insights about the current state-of-the-art and some issues at play.  

“The Internet of Things (IoT): An Introduction” (Ch. 1)

In the early 2000s, Kevin Ashton at the MIT AutoID lab was thought to have engaged in the early work that “would later become the Internet of Things”; he was looking for ways for Proctor and Gamble “to enhance its business by connecting the RFID information to the internet” (Xia, et al, 2012, as cited in Yarali, p. 1).  [“RFID” stands for “radio-frequency identification.”]  In the intervening years, connecting devices to the Internet has enabled “smart utilities, smart health, smart building, smart public services, …smart transportation,…Smart cities, energy and environmental protection, smart industries, and public safety” (p. 1).  At present, IoT is considered an area for “growth and innovation” (p. 1).  IoT is described as having five areas of operation:  smart home, smart environment, smart wearable, smart city, and smart enterprise (p. 5).  The architecture patterns vary depending on the depicted IoT contexts.  A three-tier architecture pattern involves an edge, platform, and enterprise tier with data flowing among each.  Devices connect at the edge, and data is exchanged at the platform tier (which handles the provisioning of services and which processes data), and the enterprise tier involves domain application based on defined rules (p. 6).  Another conceptualization shows how devices connect to the cloud through traditional telecommunications pipes (p. 8). In terms of industrial applications, there is the sense that it can “enhance manufacturing, energy, agriculture, transportation and other industrial departments that directly affect the economies of countries” (p. 10).  Increased awareness of the interconnectivities in human societies may inform more frugal and sustainable living.  

In this introductory chapter, Yarali addresses some basic ethics concerns around data ownership and consent, data security, and adherence to minimization approaches (only collect what is needed to provide the service) (p. 12).  He makes the case that corporations can achieve return on investment (ROI) by investing in this space because of the potential technical and social relevance.  

“Technical and Physical Aspects of IoT” (Ch. 2)

In Chapter 2, Yarali shares a diagram of aspects of technological challenges in the “technical and physical aspects of IoT”:  “security, connectivity, compatibility, standards, interoperability, intelligent analysis & actions.” (p. 19)  He takes on an empathic business point-of-view in considering device makers:  “As device makers are manufacturing IoT devices, there are inherent challenges such as time to market, interoperability, authentication, security, digital data protection and overcoming technical issues like power consumption, and limited computing process” (p. 17).  

The sharing aspects of the IoT can be hindered by any number of technological incongruencies between devices and systems; at present, the ubiquity of standards and practices for interoperability do not exist to enable “detecting, registering, activating, and correspondence abilities” (p. 19), across a range of devices.  Over a hundred IoT platforms in use are incompatible for interoperability (p. 19).  He writes, “The Intel IoT platform, the IBM ARM embed IoT device platform, ThingWox, and Axeda cloud service are examples of the various offerings of platforms based on technical or application orientations where there is none or limited regards of device connectivity to a specific industry sector” (p. 20).  The vision of interconnectivity goes like this:  

Many technology experts envision a future that is interconnected to the point where someone’s smart car can wirelessly communicate information to apartment or home, and even to medical equipment that might be in the home. This level of interconnectivity lends itself nicely to the concepts of network convergence which is a growing movement to have all types of data contained and monitored on a single type of network.  Having wearables, and smart technology that can accurately, and quickly read this information and make meaningful decisions from the information serves to make everyday life more convenient for all members of modern society.  (p. 26)  

IoT then connects people, devices, data, processes, and capabilities.  IoT may have a role in both virtual reality and augmented reality.  Messages may be targeted to particular individuals in personalized ways while protecting individual data and privacy.  Protections need to be built against the hijacking of robotics and drones through encryption (p. 28).  There need to be protections against faked data in man-in-the-middle attacks (p. 33).  Yarali posits a “minimal viable product” approach in this space:  

Given the history of the internet, the IoT will most likely not have a strict standard of architectures and protocols. Costs, prices, resources will most likely dictate which protocol will be used.  The most likely scenario will be that the lower transport protocols will be diverse and the upper protocol will become standardized and is the opposite of how the internet works now. A minimum viable product (MVP) is a product that is deployed to customers, with the minimum amount of features that are needed to gather information on how it is used so that you can fix any problems. This would be deployed to a few people for testing and would gradually increase as you perfected the product. This style of the rollout is popular with software but not so popular with hardware. (p. 29)  

For those who would build such devices, the author suggests narrowing down the list of available technologies, committing to those that would function, and building the capabilities.  He suggests that the Raspberry Pi would be a useful computer on which to build a device. (p. 34) 

“The Internet of Things (IoT) and its Important Roles in Reshaping the Environment” (Ch. 3)

The “things” in the IoT may be “a naturally occurring object, non-living object or a farm animal or a human being,” writes Yarali (2018, p. 35).  As such, IoT devices may test for air quality (p. 38).  They may inform on nature preservation endeavors using animal tracking devices, energy management through smart grids, water management (in homes, outside homes, on farms, in manufacturing, and elsewhere), and other applications. Yarali observes:  

Four fundamental device management prerequisites exist for any Internet of Things (IoT) gadget organization:  provisioning and verification, arrangement and control, checking and diagnostics, and programming updates and upkeep.  (p. 40)  

This is not once-and-done.  Device software will require bug fixes and updates.  

In terms of smart cars, human-driven cars were involved in over 1.24 million fatalities in 2010, according to World Health Organization statistics; in the U.S., around 30,000 people die from motor vehicle accidents annually (p. 41).  Properly created technologies may lower these numbers (and those of less serious accidents) drastically.  The author suggests driverless cars are on the horizon, without the need for anything with steering wheels and brakes. In terms of workplace safety, workers may be nudged to increase their levels of caution and awareness when they face dangerous contexts with smart devices (with some predictive capabilities) (p. 43).  Drones are harnessed to engage in “bio carbon engineering” (p. 56), to help in the fighting of fires and in the replanting of trees.  IoT devices may enhance human understanding of the world, such as by providing early warning for “forecasting the earthquake” and identifying resistances of snowy surfaces (pp. 56 – 57).  They may aid in the evaluation of crops and provide airborne data to inform the application of sprayed insecticides and pesticides.  In disaster context, they may enable access to spaces not otherwise accessible…and provide data that would not be available otherwise.  

The ranges of devices, if they can all interoperate with clarity and security, may enable what Yarali calls “a hyper-associated society” (p. 45).  He writes, somewhat torturously:  

A summed up structure is required for information gathering and displaying that viable abuses spatial and worldly attributes of the information, both in the detecting space and also the related changes areas.  For instance, urban clamor mapping needs a continuous gathering of commotion levels utilizing battery fueled hubs using the settled framework and participatory detecting as a key segment for well-being and personal satisfaction administrations for its tenants. (p. 50) 

The current conceptualizations of IoT leaves many points of vulnerability where data may be intercepted, changed, compromised, or misused.  While cryptography is suggested as a defense, that approach will be costly in processing and time.  

Large-scale activities “are progressing in countries like Japan, Korea, the USA and Australia, where industry, related associations, and government offices are teaming up on different projects, and similar capabilities towards an IoT by incorporating smart city activities, savvy lattice programs consolidating shrewd metering advancements and take off of rapid broadband framework.” (p. 55)  Various other engineering pieces are being worked in various countries around the world, such as advancing RFID and Wireless Sensor Networks…and setting up the Internet Protocol for Smart Objects…and co-learning about smart driving…about military applications, and others.  (p. 55)  

“The Internet of Things:  Technological, Physical, Business and Socio-Economic Environments” (Ch. 4)

The title of Chapter 4 suggests the study of a melded environment with individual sub-parts.  This suggests the importance of being aware of the interrelated parts that may affect the work. For example, advances in all the related technologies can inform IoT, such as “routing conventions, stream control power, and synchronization” in networking (p. 68).  A wide range of available data may be made available for “shrewd observing” (p. 79), but this particular chapter fails to coalesce and fails to offer unique insights.   

“The Internet of Things and 5G Wireless Networks” (Ch. 5)

5G networks in development are the hope for the IoT because the broadband services and speed (50 Mbps) enable the IoT services and data sharing with low latency.  These networks are also the focus of nation-state competition, with countries concerned about their usage by governments to compromise citizen and non-citizen privacy for government aims.  The task of building the next-generation of wireless networks with the proper “cost, flexibility, connectivity, and the use of artificial intelligence in networking” (p. 93) is no small challenge.  Mobile services “in trains, vehicles, and aircraft” is thought to be in place by 2020, which may expand IoT to less traditional locations (pp. 97 - 98).  

He also mentions the “tactile internet,” referring to “contexts where human beings control virtual and real objects wirelessly” in contexts that require “a real control audio and signal or visual feedback,” for “autonomous cars, remote medical care, and manufacturing” (p. 99), among others.  This concept offers fresh interaction modalities.  

In this chapter, Yarali raises a powerful insight:  

The transformation will occur when existing infrastructure and data can be virtualized into a true self-organizing network that is flexible and expandable that is able to meet many different types of endpoint devices and the data requirements presented by that device.  Next generation of 5G networks will become the most valuable digital asset to the economy as a whole. The digital transformation overall will change how we think about solving problems in business, life and everything in between. (p. 94)  

The vision is that data will find its way to necessary end points, and AI would use the dynamic information in ways that enable new and emergent wisdom.  The networks that are created will be self-organizing, and important groupings will emerge from the connectivity.  

“Critical Issues Facing the Internet of Things (IoT)” (Ch. 6)

A statistic referred to multiple times is that some 20 billion devices will be on the Internet by 2020, according to Gartner.  For these to function optimally in an IoT, there need to be technological standardization, secure channels for information transfer, the protection of privacy, the provision of security, the provision of data integrity, and other basics (p. 104).  Work also has to be done for long-range wi-fi through satellites.  Wired and wireless IoT both need building out.  There are so many technological requirements that advances in a number of technologies will have implications on IoT:  network coverage, location-based services, data storage, cybersecurity, GIS, and others.  

“Security and Privacy Issues on the Internet of Things” (Ch. 7)

Said another way, IoT is “simply a device with an internet connection” (p. 125).  The author cites the Shoran web index for associated gadgets online, which identified 500 million associated gadgets “including control frameworks for industrial facilities, hockey arenas, auto washes, activity lights, surveillance cameras and even an atomic plant” (p. 126). One identified challenge has been the fact that “millions of devices are running very old versions of their operating systems with many known vulnerabilities that a hacker could use to gain access to the system” (p. 126).  How people use technologies inform methods to build devices and their basic functionalities, with security testing, hardened protections, and user testing. Yarali observes (as an insider):  

With all these security controls and techniques, one would think it wouldn’t be too difficult to apply a variant of them to in the IoT world. However, in order to do that, considerable re-engineering would be required in order to address device constraints.  For example, blacklisting which is very successful on a regular network requires too much disk space to be a practical solution for IoT application.  IoT devices often have limited connectivity and are designed for low power consumption. Customarily, their processing power and memory are limited to only as much as they need to perform their task. To add to the challenge, most are ‘headless’ devices, meaning that there is no human operating them. This eliminates the possibility of someone inputting authentication credentials or decide if an application should be trusted.  In our absence, the device must decide on its own whether to accept a command or execute a task.  (p. 131)  

This author suggests methods to authenticate devices once they are logged in-to.  He proposes ways to solve issues of being aware when unauthorized changes have been made.  With the pressures of time-to-market and appealing to consumers, companies that create various IoT products still have to consider legal liabilities from security lapses. Yarali points to stolen CIA documents released on WikiLeaks to highlight the wide ranges of ways to compromise technology systems.  He cites the various layers of technologies through which compromises may occur:  the physical, the data link, the network layer, the transport layer, the session layer, the presentation layer, and the application (service) layer (p. 136).  [One useful conceptual approach is Cisco’s framework involving “Authentication, Authorization, Network Enforced Policy, and Secure Analytics.” (p. 136)  He also points to IEEE and other protocols, certification approaches, and public-key operations.  He suggests the importance of “confidentiality, integrity, and availability” (the so-called “CIA triad”) for data handling (p. 135).]  He notes that security is not achieved by any “silver bullet” but a complex range of protocols, technologies, tools, and endeavors.   

“Application and Services in IoT” (Ch. 8)

The complexities of the IoT cannot be downplayed, given the issues of scale, heterogeneous technologies and devices, ambition, multimodal data sharing, security risks, and anticipated effects on people’s lived lives (and related legal liabilities).  How technology tools are designed may motivate particular behaviors, some pro-social, and others not.  

IoT is thought to apply to different sectors of people’s lives.  In one figure, the most common uses of IoT applications, in descending order, include the following:  smart home, wearables, smart city, smart grid, industrial Internet, connected car, connected health, smart retail, smart supply chain, (and) smart farming  (2014, as cited in Yarali, 2018, p. 145).  This visual was created from a combination of social media sources including from Google searches, monthly Tweets, and LinkedIn posts.  

“Industrial Internet of Things” (IIoT)” (Ch. 9)

In terms of the promise of unique insights, Chapter 9 might include some deeper insights of applied IoT.  The author shares some IIoT applications:  enterprise resource planning (ERP), manufacturing operation management (MOM), controls, sensors, product life cycle management (PLM), customer relationship management (CRM), and supply chain management, in Figure 1 “IIot Control Areas.” (p. 159)  He describes some of the benefits of smart enterprise control:  “the large-scale production of customized products, early identification of defects in products, size reduction in product recalls, elimination of root causes by product modification, planning production alike (sic) with the weather forecast, modifying the manufacturing of products in accordance with the market value of their raw materials, etc.”  (p. 160)  In this chapter, he offers some diagrams showing an IoT business model (p. 163) and how that compares with IoT (p. 166).   

“IoT: A (sic) New Trends in Enterprise” (Ch. 10)

If academic publications are a lagging indicator about what is happening in the world, and corporations have no incentive to leak intellectual property and competitive advantage to the world, then researchers may be limited to old information and social imagery and social media posts, which results in writing that may be multiple steps away from the actual state of the field.  These challenges are quite apparent in Chapter 10.  

The author conceptualizes four phases of IoT development in business:   

1) operational efficiency in terms of “asset utilization, operational costs reduction, and worker productivity”  
2) “new products and services for “pay-per-use products, software-based services, and data monetization” 
3) outcome economy as “pay per result, newly connected ecosystems, and platform enabled marketplace” 
4) “autonomous pull economy” which consists of “continuous demand sensing, end-to-end automation, and resources optimization and waste reduction” (Gierej, 2017, as cited in Yarali, 2018, p. 172)  

“IoT: Smart Cities and Smarter Citizens” (Ch. 11)

Where this all leads is unclear, but a logical end conclusion may be one in which IoT is so pervasive that “no individual can be separated from it and still operate within the limits of a city” (p. 183).  Yarali suggests an analogy of people giving up their health information in order to receive proper care to that of citizens who give up their private information in order to engage as full citizens in a polity (p. 185).  For some, this might be ideal, but for others, this could suggest a dystopian sci-fi nightmare for people who have to accept technological predominance to live.  The author describes “non-commercial” or government-based IoT to enable city functions.  Here, there would be trade-offs between people’s personal information for public goods (from digital government). Some global cities, like Seoul, South Korea, seem to have crossed this Rubicon with their smart city.  

He suggests the advent of “personalized IoT systems” given people’s needs for privacy and their preference for selective data sharing.  In this chapter, he shifts from the third-person to the first-person to describe home automations to enable controls for temperature, security, light controls, music, news broadcasts, entertainment, cooking, and other aspects.  He describes the home he is co-building with his family and the respective technologies and linked capabilities.  He describes some of the more cutting edge designs:  a sensor that measures the amount of water coming out of a showerhead to decrease water wastage, a sensor that suggests when Febreze air freshener should be sprayed, a rain sensor to prevent water wastage from lawn watering, a wifi-connected robot floor vacuum (p. 197), and others.  Ultimately, for IoT to work, it has to create value for people.  
 
“IoT: Architecture and Virtualization” (Ch. 12)

In this chapter, Yarali proposes the use of Software Defined Networking (SDN) and Network Function Virtualization (NFV) as important approaches to enable IoT in scalable, sustainable, secure, and simple ways. (p. 201)  In terms of enabling IoT, there can be physical or virtual data centers, or some combination, with various tradeoffs.  In this chapter, he offers various general diagrams to show how the respective architectures could be set up and deployed.  

“The Internet of Things: A More Connected World” (Ch. 13)

It is thought that IoT may presage a new paradigm of human interactivity, based on real-time and mixed-time data-sharing. Perhaps the data may indicate trends from others’ experiences.  Perhaps people may engage with each other in different ways. Such potentials are emerging now.  Much of the potential is as-yet amorphous and ill-defined, and thinkers now and in the future have plenty of space to explore and create. Those that work on user interface designs, smartphone applications and hardware, cell phone technologies, high-resolution displays, “monitoring devices for healthcare and finance,” and “household appliances” (p. 205) are well positioned to contribute to current and future IoT.  Respective businesses are organizing and investing in R&D to meet the challenge.  

Some current challenges include ensuring that there is increased expansion of fixed broadband in rural and tribal areas in the U.S., which are currently at around 60% (p. 210).  Researchers are working on more stable lithium batteries.  Yarali suggests that the U.S. will not be the top player in the world in the IoT space because those with faster connectivity will come in first (p. 211).  The task is to position with the proper infrastructure, technologies, and human resources talent.  He observes that smaller countries are better able to create denser interconnectivity than those with wider geographical spaces, but there are also disparities between countries of the north-south divide.  

“Integration of IoT and AI” (Ch. 14)

In terms of cutting-edge technologies, IoT and artificial intelligence (AI) are thought to potentially converge in constructive ways. These include human-assisted AI, human-augmented AI, and autonomous AI.  “The IoT-AI amalgamation will enable intelligent application of sensors, resulting in data prediction, prescription and adaptation,” Yarali writes (p. 238), to enable real-time problem-solving and interventions (such as in aircraft flight and healthcare). IoT communication is conceptualized as involving “any moment, any location, (and) any device”(p. 239).  

A five-layer infrastructure is conceptualized, involving the sensor layer (1), the transport / network layer (2), the processing layer (3), the IoT application layer (4), and the business layer (5) (enabling “security, privacy, management, control of applications and businesses” and business analytics) (p. 241), with AI mostly affecting the two highest levels, 4 and 5 (p. 242).  The IoT-AI enables more powerful decision making (p. 244).  The author refers to multiple types of artificial intelligence:  

  • narrow or weak AI that is thought to offer the “most primary level” of AI [“processing and translating languages, identification of images, detection of face and voice, predicting the weather, etc.” (p. 245)]  [As a side note, Andrew Blum's The Weather Machine: A Journey Inside the Forecast (2019) provides a view of a complex global atmospheric model with hundreds of thousands of variables and complex data assimilations to enable fairly accurate weather modeling globally.  Blum referenced the interactive 3D Earth Nullschool resource which enables a real-time visualization of the Earth's wind patterns.  This resource could not be shown in an iframe here, so the link is provided in the prior sentence.]; 
  • general or assisted or human AI [which functions “as well as a human, with the added benefit of rapid processing and analyzing of data and offering solutions” and can generate new ideas (pp. 245 – 246)]; and 
  • super AI [in which the machines surpass human intelligence “in every possible area and even control every single system—technical, educational, etc. – that is essential to humans” (p. 246)].  



Figure 2.  Global Wind Patterns (Earth Nullschool)


 He also mentions machine learning:  

Machine Learning (ML) is a systematic procedure that involves teaching the machines to learn from the acquired and accumulated data, it is used as an application and approach where machine learning achieved by expert technicians integrating the data into the software of the machine or the device.  The Embedded data has a definite aim and parameters and is divided into instructional data and assessment data. The instructional section of the data alters the parameters to optimize the aim, while the assessment section implements the data to scenarios and tries different methods for effectiveness.  It also generates parameters to utilize the obtained knowledge in future to other unpredictable scenarios, which is advantageous, because the combination of its systematic analytical procedure and limited scenario-specific alteration capacity, enables the implementation of machine learning in problems that are understandable, and yet cannot be fathomable or easily solvable by humans… (p. 246)  

“Deep learning,” based computational neural network analyses based on bio-mimetic engineering of the human brain, may also be applied to IoT data for learning (pp. 246 - 247).   

“The Internet of Things:  Machine Learning, Artificial Intelligence, and Automation” (Ch. 15)

In Chapter 15, IoT is defined here as “a genre of services and technologies that have been created to focus on creating a network of technological objects” (p. 253) (not sure why it is defined differently across different chapters).  So how can the respective artificial intelligence and machine learning and deep learning approaches be applied to IoT?  How can these capabilities be harnessed for social goods?  What are some potential implications for 4IR, the so-called Fourth Industrial Revolution?  This chapter begins to pose questions but not to really answer these in depth.  

“The Internet of Things and IT: A Platform Transformation” (Ch. 16)

The last chapter ostensibly explores how some types of machine learning may be aligned with IoT for potential benefits but mostly summarizes some basic information about types of machine learning (supervised, semi-supervised, reinforcement learning, and automated).  He notes that machine learning cannot be applied effectively in cases where there are no human experts, when people cannot describe the tasks, when the desired outcomes of the problem solving are “variable,” and when user needs are “customized” in the problem solving context (p. 262).  His list suggests the importance of having humans in the loop.  (Said another way, machine learning is not about having machines take over human roles in analysis and decision making but more about supporting human endeavors.)  

He frames the machine learning work as involving an initial loop and an end one.  The initial step involves capturing the domain knowledge and context, selecting relevant data and processing it appropriately, interpreting the findings, and consolidating the newly learned knowledge…and then iterating in that space for new insights and with new data until the final desired results are captured (p. 266).  He describes some canonical computer science applications of artificial intelligence such as for email filtering, personalization, fraud detection, and speech recognition (pp. 270 – 271).  Of course, computational image recognition has also been achieved.  Yarali argues well for the importance of governance over this space (p. 278) along with ethical considerations based on principles and reasoning (pp. 279 – 283).  He cites a data visualization indicating the most-used AI enterprise solutions as of 2015 (in descending order): “voice recognition and response, machine learning, virtual personal assistants, decision support systems, automated responding / communications, analytics applications, robotics,” and a catch-all “all of the above” (Narrative Science 2015, as cited in Yarali, 2018, p. 289).    

New Learning about IoT?

When approaching a contemporaneous book about the Internet of Things (IoT), one may reasonably be expected to have some basic questions answered in empirical and cited ways:  

  • What is the state of the art of the Internet of Things (IoT) globally and in various respective locales?  
    • Why are some industries more advanced than others?  
    • Why are some governments more “plugged-in” than others in terms of IoT advances?  
  • What are some of the more inventive applications of IoT? Why?  In what spaces?  
    • What are some of the leading technologies in the IoT space, and why? What are the aspects of these applied technologies that may account for their success?  (And what about those that have not achieved success? And why did these IoT technologies likely not achieve success?)  
    • Who are the leading makers of the respective technologies?  
    • What are some of the ways that the movers and shakers in this space are positioning for the short-term, the middle-term, and the long-term?  Which are the bets that seem most likely to pay off?  Why?  
  • What is the state of research in IoT? Who are the main researchers?  What are some of the most important research contributions, and why?  
    • What are some theoretical and conceptual models being applied to this space?  
  • What are some of the main preferences of the human consumers of IoT devices, applications, and services, and why?  What sorts of marketing research is being done in this area, and what are some of their early findings? 
  • How is IoT being thought of as disruptors in private industry?  The public sector?  In non-profit or voluntary spaces?  
    • Where are the inroads being made at the macro levels?  
  • What are some potential game changers in this space?  The most creative applications?  
  • What are some creative uses of data collected in IoT, and why?  
  • And others 

There are some very practical and real-world ways to effectively answer some of those questions.  A researcher can invite experts to address the issue as authors or as interview subjects.  A researcher can engage in deep research in the official academic literature, in journalistic accounts, in gray literature, and in primary or first-hand research.  A researcher can explore social media for other contents that may inform on the issue.  A researcher may engage informed online professional communities and crowd-source some of the questions.  A researcher may explore IoT devices in labs, and reverse engineer software programs (when possible and when not illegal).  Such efforts will cost in effort and time, but these would be worthy of a book.  

 
 

Figure 3.  Network IoT Internet of Things (by jefferb on Pixabay) 


Or, if fresh information is not available, then there should be at least clean and effective writing to enliven the topic and provide a sense of vision and possibilities.  There should be original and engaging figures. There should be original data.  There should be original ideas, moored to the possible.  Of failing either of the two—the fresh information or the visionary writing—perhaps a project should be reconsidered, supported by the publishers, and held back from public release until the work is ready for prime time.  

Conclusion

IOT:  Platforms, Connectivity, Applications and Services (2018) reads like a work rushed to publication, without sufficiently new insights, and with the sense that this might have been written for a particular course for the author. Many of the observations seem pedestrian, in part because of the wholesale integration of others’ informational graphics into the work (from companies, from organizations).  In many instances, figures seem to be captured from various commercial sources without apparent rights releases or formal citations. When some in-text source citations are included, they are a mix of different types (some permutation of APA and MLA and end note).  [There is no consistent quality in the visuals either. Closer following of figure conventions and labeling…and the creation of original figures…would have strengthened this work greatly.  On p. 149, the informational graphic has a misspelling of “jewellry” (sic).  Third-party visuals from companies and organizations may be blurry, with stretched aspect ratios, and missing citations except for residual text data on the images.]  Consistently, there are missing definitions, awkward sectional and paragraph transitions (or missing ones), missing research source citations, poor transitions, incorrect grammar and syntax (such as parallel phrasing), and typographical errors. On pp. 84 – 85, the author apparently refers to contents on a web page, an “IoT Historical Timeline page here” and an “IoT Overview Infographic,” which begs the question of whether this book is a result of contents initially on a website; regardless, such citations to nowhere are sloppy and unprofessional. Then, Ray Bradbury, author of the short story, “There Will Come Soft Rains,” is misidentified as “Ray Bradburn” (p. 189).  Who does that?  (Such problems reflect on the author, the university, and the publisher.)  

In rare cases, it is possible to slog through poor writing (with a high density of mistakes) if there are original ideas or fresh insights, but not so in the absence.  For example, the Preface opens: “Technology has been defined to have a significant reputation that has been highly emulated in the world of production” (Yarali, 2018, p. ix).  Huh?  One chapter refers to a “flag scanty” (p. 51), which does not seem to result in any hits on a Google Search.  (Witty and smart technology terms that are novel are always welcome, but not free-floating made-up words or typos.)  One agreement problem is in a title:  “IoT:  A New Trends in Enterprise” (Ch. 10). In parts of the text, the point-of-view shifts from the third-person POV to the first-person one.  Readers can expect better.  Clearer fact-checking, academic standards, citations, and close-in expertise would strengthen this work.  It is also not clear who the audience is for this work—mainstream readers or technologists, or those who are both, and others interested in this topic.   

The work reads somewhat abstractly, as if the author had parachuted into the topic, read some magazine pieces about the topic, and offered some general “best practices” from disparate information technology engineering contexts. The technological and engineering framework would not be a problem if the writing were more coherent in some parts.  (Along with a strong editorial hand, a simple grammar and spell check would help.)  In authored texts, the table of contents should have a clear trajectory and organizational structure.  However, the 16 chapters in this authored text do not offer a coherent sequence (except a general move from the simple to the slightly more complex), and the contents within each repeat in some areas and leave gaping content holes in others. In this work, assertions and claims are made without in-text citation. Writing to an outline format (with the lettering and numbering at the respective levels and sub-levels) may be another way to help structure the work.  A critique of the manuscript should be at the macro or manuscript level as well as at the level of sections, chapters, and sub-parts. This book could have been collapsed into one long nonfictional essay or a short series and been effective.  

The related technologies and practices have been around for decades (often in the background), but IoT seems to still be on the incline of the initial curve of Gartner’s Hype Cycle.  The sense of potential of IoT is manifold, with a potential to serve as a panacea for various ills in the world (p. ix).  The inventions are dynamic and on-going, and these factors may explain the general level of specificity of this text.  

Finally, for all the critiques, IOT:  Platforms, Connectivity, Applications and Services (2018) does address a timely issue with widespread interest.  It does generally relate to in-world technologies, instead of any magical thinking or fairytale version.  It hones to economic and technological practices, generally speaking.  The general ethical practices align with contemporaneous values, including the idea that the IoT is for the world, not particular nationalities or languages alone (p. 145).  There is a sense among those in industry that there are many unexplored and unexploited potential advantages.  In many spaces, the sense is, “How can these technologies be harnessed here, for increased efficiencies and value-added features?” 








About the Author

Shalin Hai-Jew works as an instructional designer at Kansas State University. Her email is shalin@k-state.edu.  


Comment on this page
 

Discussion of "Book Review: Understanding the IoT as Interdependent Technologies"

Add your voice to this discussion.

Checking your signed in status ...

Previous page on path Cover, page 19 of 23 Next page on path