CHAPTER 5

The Internet of Things

Understanding the IOT

The tech industry is often guilty of pushing technology solutions to consumers without focusing on the benefits, the emotions and simplicity. The IOT, on the other hand, is a pithy catchphrase, but by itself, it means very little when it comes to actually buying, implementing, or solving something. The IOT is a neologism coined by Kevin Ashton¹ to encapsulate a number of related concepts. It involves smart and connected products, multiple types of open and closed networks, robotics, cloud-based access, decision analytics, and functions ranging from monitoring to control and optimization. Many people also refer to M2M communications, which are distinct from the content and functionality that is designed for consumption by people, or even people interacting with machines.

At its heart, it involves allowing everyday objects or things to be connected to the Internet via sensors and accessible via the Internet. These could be pipes managed by utility companies or buoys in the ocean, but are increasingly consumer-grade devices as well. The sensors themselves are also varied—based on what it is that they are sensing. Pressure sensors on a utility pipe could help ensure smooth water flows, or track leaks, whereas motion sensors on buoys can provide early warnings of tsunamis. The data from these sensors may be public or private but can be married with other data to drive outcomes for users. The data may also be read by systems to automate decisions based on preset limits. For example, if the buoys show that the waves are dangerously high, the system can send out a message to vessels in the area.

The IOT therefore includes communication between machines, between people and machines, and also between people and people via machines. It includes wearables, and all manners of sensors, and an ever-increasing ocean of data, and an implicit assumption of an economically viable, reliable, and available network.

Microelectromechanical sensors (MEMS), which are smaller than a millimeter in size, are a very popular type of sensor. They go into a lot of consumer products—including smartphones and inkjet printers. They are also found in health care devices, such as stents. The launch of the iPhone created a spike in demand of 10 million MEMS in 2007, but the MEMS market grew to 15 billion in 2015! You can almost pick any number for the global sensor market in five years, and it might turn out to be a conservative estimate. One estimate quotes 100 billion sensors by 2025, with a revenue of $10 trillion. You’re probably already using more sensors that you know, and you’re definitely consuming data from more sensors than you think.

Michael Porter and James Heppelman in their very lucid HBR article posit² that all products in future should have mechanical/electrical components, but also software components and communication components. These three collectively make products smarter and ultimately evolve to product systems (e.g., home security) and then to a system of systems model (e.g., connected homes)—which spans an entire problem domain, according to the authors. The kind of activities that we can perform on smart products evolves from monitoring, to control, optimization, and then to autonomy. Ultimately, this leads to improved competitive performance via operational efficiencies and strategic positioning choices. Often, forcing the question “what business are we in?”

Internet of Your Things

Try this exercise. List all the things you interact with. From your house and home to the trains you take and from the clothes you wear to the hotel room you might live in on my work trip or holiday. You’ll quickly realize that the list is too long to actually complete. But also, you might see that this is a hierarchical list. Your home, for example, is a complex construct, and comprises many sub-things. For example, rooms, walls, plumbing. Some of these, such as heating may have further subcomponents—radiators, boilers, and so on. You can then try and imagine how you would want each of these things to behave to make your life easier. I would like my wardrobe to be aware of my calendar and the weather and recommend to me choices for what to wear with visual combinations of clothes to wear every morning. You might want your coffee machine to sense when you’re up and about and start brewing your favorite cup of coffee, and remind you when it’s ready.

Let’s consider just one of these things—windows. Everybody has windows at home, and they affect our everyday lives. They have states (open/shut), based on the environment and conditions. We typically associate safety, air conditioning, and sunlight with windows being open or closed, based on the weather, time of day, and so on. You can construct a table in your mind with benefits and the associate state or activity for those benefits, and the conditions under which they need to be activated. For example, when it rains, your windows need to be closed so that your house can be dry. During a pleasant and clear day with lots of sunshine, your windows need to be open so that your house can be aired. At night, they need to be closed so that you can be safe. Perhaps they can change their level of transparency in response to the amount of sunlight.

You would be right in thinking that all that we’ve described above can happen with smart windows, which need not be connected to the Internet. We need to distinguish between smart (computation capability) and connected (communication capability), and not treat them as synonyms. Some windows come with a remote control, they can be opened and closed, and they can also react to weather conditions and close if left open when it starts to rain. They are actually smart, in some way, and possess limited capability to communicate. They’re just not on the Internet. The challenge of this model is that your ability to control these outcomes is limited to the preset automations and your being in close proximity—that is at home, and with the right remote.

Some of my home windows have rain sensors, which allow them to automatically close if it starts to rain, but they don’t have a sun sensor, which allows them to reopen when the sun comes out again. Actually, for all my windows, I would like a certain level of intelligence. I’d like to be reminded if ground floor windows are left open at night or when I’m away. If I had pollen allergies, I would probably like to be alerted if the pollen count is too high, or have the windows close. I would like to be able to open all multiple windows or close them, even if I’m not at home, based on weather conditions. I might even have settings for sunny day, which applies a set of commands to all windows. This is the optimization that we ultimately want to get to. These controls should extend to blinds because they are a part of my window product systems, whereas currently, we might have completely different suppliers for these two products (windows and curtains/blinds). Any maker of smart windows must therefore consider blinds and curtains as a part of their product system.

Now, considering any smart and connected product, we could argue that they have sensors, which generate data, which are used by apps, which enable access and control of the product, and provide additional functions that ultimately deliver a benefit. The sensors are obviously on-board the device/product. But the data generated could be anywhere, typically on a cloud, so that the apps and the access can take place through any connected control point (such as a mobile phone).

This is where the IOT really kicks in. In my window example, for the Velux models that we have installed, the data, access, applications, and controls all sit within a closed system involving the window and the remote control. A mature IOT model requires a cloud-based data and access model and an ability to use the data and control/monitor the product from any device and application that is authorized. So, I could then control the windows from office if I wanted to air the living room on a sunny day.

The Network of Your Things

Many smart products today are capable of performing advanced functions autonomously, which have nothing to do with the Internet of anything. The Roomba smart vacuum cleaner is a great example of an exceptional product that doesn’t really need to connect to the Internet for its everyday operation of intelligent vacuuming.

Most individual products also tend to ignore or be indifferent to the network effect, which kicks in when we consider multiple elements in the same network. My windows may be rain-sensitive, but I might have other devices, products, and appliances at home, which may be influenced by the occurrence of rain. Does each product need to have its own rain sensor? In my IOT wish list, my smart windows can communicate to other appliances at home. The washing machine can run an extra spin cycle when it rains, so clothes dry in the same time, and conversely, when it’s sunny, it can reduce the spin cycle to conserve energy. This is when we get to the real IOT. Today, the way we connect devices is actually via the cloud. Each device connects to its cloud and data is shared across providers at the backend.

B2B and M2M

A significant part of the IOT will continue to involve machines talking to machines behind the scenes in ways that will not involve humans at all. This may well be the lion’s share of the IOT in the years to come. This could be cars, trains, buses, and transport infrastructure communicating with each other, or your devices sending data back to manufacturers systems, or even production systems and factory infrastructure exchanging information—all of which will be used for decision making that will go on without us as humans being explicitly involved. A simple example that has been discussed often of late is you run your dishwasher at night after dinner, but your smart dishwasher talks to your smart meter and runs the cycle after midnight, when the energy cost is at its lowest. The communication, data exchange, and decision making are all happening without your explicit involvement.

Challenges

I’ve made the point more than once about how easy the browser-based Web world was. In the IOT world, though, we have to think about the hardware, the embedded system, the interface and communication protocol, the environment, and purpose—almost everything has to be designed from scratch. What’s the user interface of a thing? If it’s a sensor in a coffee machine versus in a door, how should we access the data, how can interact with the thing? The design challenge moves from the interface design to experience and even environment design, and it will vary from fitting an antenna while managing heat dissipation, to figuring out how to retain product aesthetics and usability while adding sensors to running shoes. Service design, which we discuss later, is fundamental to the creation of IOT models. We must take a design centric view and build from there. That’s the only way we’ll get around to focusing on the right problems to solve, to ensure adoption.

Ultimately, this will create and shift value, destroy old models, and create entirely new services. We usually find it easier to think of things that will be better, rather than different. A common example of better, is that a future fridge will tell us when it’s out of milk, or order it on our behalf. The fridge is still doing the same basic job. An example of different might be that our fridge will become a health care device and do an analysis of all the food we’ve consumed over a month, and give us a health trend report, or it will point to foods that are missing from our consumption basket, or even calculate the environmental footprint of all the food we eat.

Delivering Value in IOT

The IOT is one more interface for businesses, generating streams of data from physical environments. The IOT offers great fortunes for companies who will be bold enough to rethink their business models and honestly answer the question “what business are we in?”—allowing them to move from selling a product to delivering a composite service, which may include a physical product. It might even mean changing the commercial model where the product is only leased to the consumer who actually buys the service rather than acquire an asset.

Undoubtedly, the IOT is a big deal. We’re talking about billions of connected devices changing the way we live our everyday lives. The transformative potential of this can barely be imagined. This will make a big difference to the way we deal with the sustainability, caring for an aging population, or getting supplies more efficiently to the needy, across the world. Although, there may be ongoing debate about whether our kettle should gossip with our washing machine! Meanwhile, I continue to dream about smart, connected windows, which can deliver safety, sunshine, and comfort to my home. As far as consumers are concerned, the I in IOT should really stand for invisible technology.