Part I – IoT Application Domains and Case Studies

Part I of this book focuses on selected IoT application domains and provides detailed illustrative use cases for each one. It is virtually impossible to provide a complete and holistic overview of all possible applications for the IoT, because the IoT will affect nearly every aspect of our lives; application use cases are found within the areas of healthcare, smart homes, smart buildings, smart cities, energy, agriculture, transportation and connected vehicles, the military, logistics and supply chain, retail and wholesale, manufacturing, mining, and so on. The opportunities seem unlimited, and many application domains either overlap or have strong interdependencies. Machina Research’s approach to mapping out the different applications and their overlaps on a high level is presented in the diagram below. From a forecasting perspective, this already represents a detailed taxonomy of IoT use cases, and the underlying forecasts by Machina Research contain an even greater level of detail, spanning 223 different applications and market segments.

Machina Research’s taxonomy for the IoT

While this is an appropriate level of detail for forecasting our connected future world, the reality is likely to be far more fragmented. Even relatively standard and well-defined IoT applications can be implemented in radically different ways. Consider the example of smart electricity metering. This could be supported by cellular, power-line, radio mesh, or Low Power Wide Area (LPWA) communications. Indeed, as we will see later, a single contract may include more than one of these technologies. But the fragmentation extends far beyond even this level, and includes treatment of the data provided by the smart meter (depending on whether readings are provided to a utility only, or to an in-home consumer device also), as well as the degree to which a smart meter can control in-home smart energy-consuming devices, and in what way. In some countries, the capability to support pre-pay options may be required, while in others, regulators are encouraging multiple utilities (electricity, gas, water, etc.) to “share” a single connection with the aim of increased efficiency. In some markets, integration with low-voltage generators (i.e. local or micro-generation) is a particular requirement. The list of technical and process-related factors by which two smart electricity metering solutions may differ is itself almost endless. And that’s before we consider the list of commercial and contractual parameters that may differ between solutions, leading to differences in data sharing and management, support infrastructure, and even technical infrastructure given the desire to optimize risks and returns.

The conclusion, however, is clear: while the taxonomy above might be appropriate for forecasting IoT markets, the reality is that IoT project managers will be confronted with a great deal more complexity and fragmentation than this taxonomy even hints at.

The majority of this section of the book will focus on a detailed analysis of application domains and case studies. Before moving on to that discussion, however, it may be helpful to briefly discuss some of the forecast figures associated with our connected future. In terms of connected devices today, of course, PCs, tablets, and mobile handsets are by far the dominant category. These devices are almost ubiquitous in the developed world, and their adoption rates in the developing world are rising fast. As is often observed, the real growth in connected devices in the coming years will come from connecting “things”, not people – hence the term “Internet of Things”.

It is clear that some domains will give rise to a much greater number of connected devices than others. Fundamentally, this is a consequence of the size of addressable markets. In the case of connected consumer electronics, HVAC solutions, and building security, the number of connections can be analyzed at the level of individual households: a large number of households can be expected to have several of these connections. Figures for applications like connected vehicles and smart meters can also be estimated at 1-2 per adopting household. However, applications like connected ambulances will always be far fewer in number. That’s not to say that there is no value in connecting ambulances – in fact, connecting an ambulance to a hospital so that patients’ vital signs can be communicated is incredibly valuable – it’s just that there will never be many connected ambulances, as the incidence rate of ambulances in developed economies seems to be roughly 1 per 10,000 population.

Another interesting observation is that particularly high-volume applications tend to be relatively homogeneous. Not only would a manufacturer of connected TVs expect to sell a large number of any one particular model, but one connected TV isn’t really all that different from another. Conversely, low-volume industrial applications can be very diverse: there is a world of difference between a connected ambulance and a monitoring solution for ammonia (fertilizer) tanks for farms.

The following figure highlights the forecast number of IoT device connections in certain key domains.

Connected devices chart


Instead of trying to cover all possible applications in all of these domains, we decided to delve a little deeper into a subset of them. Our selection was based on accessibility and the domain expertise of the authors, but also took into account growth projections and the perceived importance of particular applications in terms of their potential to define the pace of developments in the IoT space as a whole.

Naturally, these case studies represent different levels of maturity and innovation. Some are highly mature and widely deployed IoT solutions (especially those with their roots in M2M), while others are more innovative, less mature projects and pilots. For each case study, we tried to understand the problem domain, the specific problem, the way in which IoT (or sometimes M2M) helped to address the problem, and the lessons to be learned.

It was particularly important to us to collect the lessons learned and best practices drawn from these case studies in a structured way, because these form the basis for the Ignite | IoT Methodology presented in Part II. While working on the case studies and developing the methodology, we identified some useful tools for analyzing and designing IoT solutions. As you will see, we have tried to present as consistent a picture as possible by using the methodology tools from Part II to document the case studies in Part I. This represents a bit of a chicken-and-egg situation, as we will make use of certain tools in Part I before they are formally introduced in Part II. However, we believe that most examples are fairly self-explanatory, and help to make the case studies from all these different domains more accessible and easier for the reader to compare. If need be, you can always take a sneak peek at Part II – take a look at tools like the Asset Integration Architecture (AIA), for instance, which is introduced in the section on Architecture Blueprints.

The application domains we will focus on include:

  • Smart energy, including generation, metering, and Cross-Energy Management (CEM)
  • Manufacturing and industry, including some unconventional case studies like the Large Hadron Collider at CERN
  • The automotive sector, in particular connected vehicles
  • Smart cities, including a detailed case study of the city of Monaco