The first question that we should attempt to answer is, of course, what is the Internet of Things? Although the concepts we call on throughout this article are relatively straightforward, people have many different visions of what the phrase means, and many of the implications are hard to grasp. So we will take this question slowly in this chapter and look at it from a number of different angles.
What does the phrase “Internet of Things” mean? And how does it relate to the earlier buzzword “ubiquitous computing”? For those who are interested in the history of technological progress, where does the Internet of Things sit in the broad sweep of things, and why are we talking about it now? For those who understand best through metaphors, we look at the idea of enchanted objects, an image that has described technology for millennia but which is especially potent when describing the Internet of Things. For the more practical readers who understand by seeing examples of real things, we sketch out some of the exciting projects that give a good flavour of this exciting field. Let’s start with this last approach, with a short piece of “design fiction”.
The Flavour of the Internet of Things
The alarm rings. As you open your eyes blearily, you see that it’s five minutes later than your usual wake-up time. The clock has checked the train times online, and your train must be delayed, so it lets you sleep in a little longer. (See http://makezine.com/magazine/make-11/my-train-schedule-alarm-clock/.)
In your kitchen, a blinking light reminds you it’s time to take your tablets. If you forget, the medicine bottle cap goes online and emails your doctor to let her know. (See http://www.vitality.net/glowcaps.html.)
On your way out of the house, you catch a glow in the corner of your eye. Your umbrella handle is lit up, which means that it has checked the BBC weather reports and predicts rain. You sigh and pick it up. (See http://www.materious.com/#/projects/forecast/.)
As you pass the bus stop on the way to the station, you notice the large LCD display flash that the number 23 is due. It arrives when you turn the next corner. When the bus company first installed those displays, they ran on the expected timetable information only, but now that every bus has GPS tracking its location, they simply connect to the bus company’s online service and always give the updated information. Various transport organizations have implemented this. London’s TfL has some useful information on their signs at http://www.tfl.gov.uk/corporate/projectsandschemes/11560.aspx.
When you get to the station, your phone checks you in automatically to a location-based service (such as Foursquare). On your mantelpiece at home, an ornament with a dial notices the change and starts to turn so that the text on it points to the word “Travelling”. Your family will also see later that you’ve arrived at “Work” safely. (See http://wheredial.com.)
On your lunch break, a pedometer in your training shoes and a heart monitor in your wrist band help track your run around the block. The wrist band’s large display also makes it easy to glance down and see how fast you are running and how many calories you’ve burned. All the data is automatically uploaded to your sports tracking site, which also integrates with your online supermarket shopping account to make it easy to compare with how many calories you’ve eaten. (See http://nikeplus.nike.com/plus/.)
As you can see from the preceding links, each of these products is feasible with today’s technology. Each has been prototyped, and many of them exist as craft or mass-market products.
The “Internet” of “Things”
We’ve looked at a number of examples of the Internet of Things, so what is the common thread that binds them together? And why the name? All the cases we saw used the Internet to send, receive, or communicate information. And in each case, the gadget that was connected to the Internet wasn’t a computer, tablet, or mobile phone but an object, a Thing. These Things are designed for a purpose: the umbrella has a retractable canopy and a handle to hold it. A bus display has to be readable to public transport users, including the elderly and partially sighted, and be able to survive poor weather conditions and the risk of vandalism. The sports bracelet is easy to wear while running, has a display that is large enough and bright enough to read even when you are moving and will survive heat, cold, sweat, and rain.
Many of the use cases could be fulfilled, and often are, by general-purpose computers. Although we don’t carry a desktop PC around with us, many people do carry a laptop or tablet. More to the point, in almost every country now, most people do carry a mobile phone, and in many cases, this is a smartphone that easily has enough power for any task one could throw at a computer. Let’s see how well one could replicate these tasks with a smartphone.
Viewing your bus provider’s timetable with a smartphone web browser seems to fulfill the same function at first glance. But just consider that last phrase, “at first glance”. On arriving at the bus stop, one can simply glance at the computerized timetable and see when the next bus is due. With a smartphone, if you have one and can afford the data use (which may be prohibitive if you are a foreign tourist), you have to take the phone out of your pocket or bag, unlock it, navigate to the right website (this may be the slowest and most complicated part of the process, whether you have to type the URL or use a QR code), and read the data from a small screen. At this time, you are not able to fully concentrate on the arriving buses and might even miss yours.
You can track your runs with an app on your smartphone, and many people do: the phone has GPS, many other useful sensors, processing power, an Internet connection, and a great screen. But it turns out that such a phone isn’t easy to carry on a run without worrying about dropping it or getting it wet. Plenty of carrying options are available, from a waist bag to an arm strap. The latter, in theory, enables you to read the device while you are running, but in practice reading details on the screen can be hard while you are jiggling up and down! To get around this difficulty, apps such as RunKeeper provide regular audio summaries which can be useful (www.runkeeper.com). Ultimately, a phone is a perfectly capable device for tracking your run, and most runners will find it a sufficient, comfortable, and fun way of logging their running data. However, others may well prefer a device worn as a watch or wristband, designed to be read on the move, worn in the rain, and connected to peripherals such as heart monitors.
Of course, no mobile phone (or even a tablet or laptop) is large enough or waterproof enough to use as an umbrella. However, you could pair a smartphone with a normal “dumb” umbrella, by checking an app to see whether it is likely to rain later before you leave the house. Unlike a calm, subtle light in the umbrella stand, glimpsed from the corner of your eye as an ambient piece of information to process subconsciously when you pass it on the way out of your home, an app requires you to perform several actions. If you are able to establish and maintain the habit of doing this check, it will be just as effective. Rather than having greater capabilities, the smart umbrella simply moves the same intelligence into your environment so that you don’t have to change your routine.
So the idea of the Internet of Things suggests that rather than having a small number of very powerful computing devices in your life (laptop, tablet, phone, music player), you might have a large number of devices which are perhaps less powerful (umbrella, bracelet, mirror, fridge, shoes). An earlier buzzword for roughly the same concept was “ubiquitous computing”, also known by the ugly portmanteau “ubicomp”, and this also reflects the huge number of possible objects that might contain computing technology. Now that the Internet is a central pipe for data, it’s hard to imagine, for example, a PC that doesn’t have an always-on broadband connection. Younger readers may never have seen such a thing. As a technologist and columnist Russell Davies joked at the 2012 Open Internet of Things Assembly in London:
I can’t understand why teddy bears did not have wifi before. A bear without wifi is barely alive, a semi-bear.
The definition of ubicomp, however, would also include the Glade air fresheners which release scent when they detect movement in the room as part of its domain. That is to say, such a device is an intelligently programmed computer processor, driven by sensors in the real world, and driving output in the real world, all embedded into an everyday object. These factors make this ubicomp, and it is only differentiated from the “Internet of Things” by the fact that these days most of the really interesting things done with computing also involve an Internet connection.
But what does it mean to “connect an object to the Internet”? Clearly, sticking an Ethernet socket into a chair or a 3G modem into a sewing machine doesn’t suddenly imbue the object with mysterious properties. Rather, there has to be some flow of information that connects the defining characteristics of the Thing with the world of data and processing represented by the Internet.
The Thing is present, physically in the real world, in your home, your work, your car, or worn around your body. This means that it can receive inputs from your world and transform those into data which is sent onto the Internet for collection and processing. So your chair might collect information about how often you sit on it and for how long, while the sewing machine reports how much thread it has left and how many stitches it has sewn. In subsequent chapters, we talk a lot about “sensors”.
The presence of the Thing also means that it can produce outputs into your world with what we call “actuators”. Some of these outputs could be triggered by data that has been collected and processed on the Internet. So your chair might vibrate to tell you that you have received the email.
We could summarize these components in the following appealingly simple (though, of course, also simplistic) equation:
Physical Object + Controller, Sensor, and Actuators + Internet = IoT
Note that in all the cases we’ve looked at, the form of the object follows the function of the Thing: your chair is designed to sit on, the sewing machine to sew at, and so on. The fact of also being connected to the Internet and having general-purpose computing capabilities doesn’t necessarily have an impact on the form of the object at all. (One might argue that current-generation smartphones and tablets are in forms optimised for use as general-purpose computers, not as portable telephony devices. Certainly, on seeing the number of phones with scratched screens, one could ask whether they are designed to be easy to hold securely and resistant to drops and the impacts of everyday use.)
The Technology of the Internet of Things
In starting to define the Internet of Things, we compared it to the earlier concept of ubiquitous computing. We could compare that, in turn, with Bill Gates’s famous vision in 1977 of “a computer on every desk and in every home” (http://danbricklin.com/log/billg_entwof.htm) and again with the earlier notion of a computer as an astonishingly expensive and specialized machine, accessible only to universities, some forward-thinking global corporations, and the military. It is worth taking a little time to look at the Internet of Things through a lens of the history of technology to more clearly understand how and where it fits.
Technology’s great drivers have initially been fundamental needs, such as food and water, warmth, safety, and health. Hunting and foraging, fire, building and fortifications, and medicine grow out of these needs. Then, because resources for these things are not always distributed where and when one might like, technological advances progress with enabling and controlling the movement of people, their possessions, livestock, and other resources. Trade develops as a movement of goods from a place where they are plentiful and cheap to one where they are rare and valuable. Storage is a form of movement in time—for example, from harvest time, when food is plentiful and cheap, to the following winter, when it is highly valued.
Information becomes key, too—hence, the development of language to communicate technology to others. Travelers might pass on messages as well as goods and services, and an oral tradition allows this information to pass through time as well as space. The invention of writing makes this communication ever more important and allows, to some extent, human lives to be preserved in words by and about writers, from the ancient philosophers and poets to the present day. From writing, via the telegraph, radio, and television, to digital information, more and more technology has been about enabling the movement of information or doing interesting things with that information.
But the other human needs we looked at haven’t ceased to exist, nor will they. We still need to eat and drink. We still need light and warmth. We still need love and friendship. We still need chairs, clothes, and shoes; means of transport and communication; and ways to entertain ourselves. The shape and details of all of these things will change but not the needs they address.
As technology has progressed, new categories of objects have been created: in the electronic age, they have included telephones, radios, televisions, computers, and smartphones. As with most new technology, these devices tended to start out very expensive and gradually come down in price. Demand drives down prices, and research leads to optimization and miniaturisation. Ultimately, it becomes not just possible but also feasible to include functionality that would previously have required its own dedicated device inside another one. So although a television screen would originally have physically dominated a living room, not only are today’s flat-screen panels more compact, but the technology is so ubiquitous that a high-resolution screen capable of displaying television content can be embedded into a door frame or a kitchen unit, and of course, even smaller screens can find their way into music players and mobile phones.
Similarly, with computers, it has become so cheap to produce a general-purpose microchip in devices that your washing machine may contain a computer running Linux, the cash register at the supermarket may run on Windows, and your video player may run a version of Apple’s OS X. But as we’ve already hinted at, mere computing power isn’t a sufficient precondition for the Internet of Things. Rather, we are looking at computing power linked on the one hand to electronic sensors and actuators which interact with the real world and on the other to the Internet. It turns out that the rapid sharing and processing of information with services or other consumers is a huge differentiator.
As an example, let’s consider the computers that exist in modern cars: they have myriad sensors to determine how well the car is running—from oil gauge and tire pressure to the internals of your engine. As well as diagnostics, computerized brakes may assist the driver when the processor spots conditions such as the wheels locking or spinning out of control. All this is local information, and although the processing and analysis of this data may be highly sophisticated, it will be limited to whatever your car manufacturer has programmed. But perhaps your car also tracks your location using GPS: this is external (although not necessarily Internet-related) data. High-end cars may communicate the location back to a tracking service for insurance and anti-theft purposes. At this point, the car carries computing equipment that is able to not just passively consume data but also to have a dialogue with an external service. When your car’s computer is connected to the Internet (regularly or permanently), it enables services such as responding to traffic conditions in real-time by rerouting around them. Your GPS might already supply such data, but now it can be created in real-time by “social route planning” based on the data aggregated from what other connected drivers nearby are doing. When the previously internal data gets connected to the Internet, the ways it can be processed, analyzed, aggregated, and remixed with other data open up all the possibilities that we’ve seen in existing connected areas and indeed new ones that we can’t yet imagine.
So there is a real change to an object or appliance when you embed computing power into it and another real change when you connect that power to the Internet. It is worth looking at why this latter change is happening now.
When the Internet moved out of academia and the military, with the first commercial Internet service providers (ISPs) opening for business in the late 1980s, the early adopters of the consumer Internet may have first gone online with a computer running an Intel 486 chip, costing around £1500, or around the price of a small car. Today a microchip with equivalent power might set you back around £0.50, or the price of a chocolate bar. The rapid rise of processing power, and the consequent cost decrease, is not a new insight: it is widely known as Moore’s law (the rule of thumb, suggested by the co-founder of Intel, that says the number of transistors you can fit on a silicon chip will double every 18 months).
However, the kind of price difference we’ve mentioned isn’t merely a question of degree: it is a qualitative as well as a quantitative change. This is a “long tail” phenomenon through which we have now hit the right price/performance sweet spot that means the cost of the computing horsepower required to talk to the Internet has fallen to a level where adding a network or computing capability is akin to choosing what type of material or finish to use—for example, whether to use a slightly more expensive wood veneer. Either option would add a little to the cost of the product but could also add disproportionately to its value to the customer. When Internet Capable computing cost thousands of pounds, this wasn’t an option, but now that it costs tens of pence, it is.
So the price of computing power has come down to affordable levels, but this is only part of the story. Manufacturers of electronic products have started to incorporate general-purpose computer CPUs into their products, from washing machines to cars, as they have seen that it has become, in many cases, cheaper to do this than to create custom chips. The wealth of programming and debugging resources available for these platforms has made them attractive to hobbyists and the prototyping market, leading to the proliferation of the microcontrollers.
Internet connectivity is also cheaper and more convenient than it used to be. Whereas in the past, we were tied to expensive and slow dial-up connections, nowadays in the UK, 76% of adults have broadband subscriptions, providing always-on connectivity to the Net. Wired Ethernet provides a fairly plug-and-play networking experience, but most home routers today also offer WiFi, which removes the need for running cables everywhere. While having an Internet-accessible computer in a fixed location was useful to those who needed to use it for work or studies, it would often be monopolised disproportionately by male and younger members of the family for general browsing or gaming. Now that the whole family can go online in the comfort of the living room sofa or their own room, they tend to do so in greater numbers and with ever greater confidence.
For situations in which a fixed network connection isn’t readily available, mobile phone connectivity is widespread. Because the demand for connectivity is so great now, even embryonic solutions such as the whitespace network are available to use the airspace from the old analog TV networks to fill gaps.
Another factor at play is the maturity of online platforms. Whereas early web apps were designed to be used only from a web browser, the much-heralded “Web 2.0”, as well as bringing us “rich web apps”, popularised a style of programming using an Application Programming Interface (API), which allows other programs, rather than just users, to interact with and use the services on offer. This provides a ready ecosystem for other websites to “mash-up” a number of services into something new, enables mobile phone “Apps”, and now makes it easy for connected devices to consume.
As the online services mature, so too do the tools used to build and scale them. Web services frameworks such as Python and Django or Ruby on Rails allow easy prototyping of the online component. Similarly, cloud services such as Amazon Web Services mean that such solutions can scale easily with use as they become more popular.
The best known of Arthur C. Clarke’s “three laws of prediction” states
Any sufficiently advanced technology is indistinguishable from magic. —http://en.wikipedia.org/wiki/Clarke’s_three_laws
We’ve already seen how technology has evolved to meet our needs and desires. The parallel invention of magic serves largely similar goals. After all, the objects in folktales and fairy tales are often wish-fulfillment fantasies to fill the deepest desires: if only I had enough to eat; if only my mother was well again; if only I could talk to my friend even though I’m far away; if only I could get home; if only I didn’t have to work every hour of the day to earn enough money for my family to eat. Literary and anthropological scholars have long studied fairy tales for the lessons that can be learned about the basic rules of human narrative and meaning and have analyzed the characters, storylines, and objects found within them. For example, the formalist scholar Vladimir Propp categorized the folktales of his native Russia and categorized their plot elements into 31 functions, including “violation of interdiction”, “villainy”, “receipt of a magical agent”, “difficult task”, and so on.
More recently, and from the point of view of a Silicon Valley entrepreneur and technologist, David Rose has talked about Enchanted Objects at TEDx Berkeley (http://tedxtalks.ted.com/video/TEDxBerkeley-David-Rose-Enchant) and has categorized various objects drawn from fairy tales and fantasy literature in ways that apply as much to technological objects. For Protection, just as magical swords and helmets protected the protagonists of fairy tales from their enemies, so has much of the development of science and technology throughout history been driven by the need for military superiority, for the purpose of security or conquest. Health has been a driver for many quests to find an ingredient for a health potion and for research into various branches of medicine, pharmacology and surgery, physiotherapy, and diet. Humans have always desired Omniscience, from Snow White’s wicked stepmother asking “Mirror mirror on the wall, who’s the fairest of them all?” to the friends settling an argument of fact by looking up articles from Wikipedia on their smartphones. Human Connection, even when one’s loved ones are far away, is an urgent, aching need: the Finnish hero Lemminkäinen’s family know that he has been hurt when the enchanted comb that he left on the mantelpiece starts to bleed. Similarly, the postal service, telephones, and social networking help keep us in touch with our family and friends. The ancient storytellers yearning for Effortless Mobility invented seven-league boots, flying carpets, and even teleportation. Through technology, we have invented cars and railways, bicycles, and airplanes. The need for Creative Expression is fulfilled in stories by the enchanted paintbrushes or magic flutes and harps, while we have always used technology to devise such creative outlets, from charcoal to paint to computer graphics, or from drums to violins and electronic synthesizers.
So, technology has always been associated with magic, and so this will be true almost by default for the Internet of Things. But there is more to it than that: a key element of many enchanted objects is that above and beyond their practical enchantment they are given a name and a personality—implying an intelligence greater than strictly necessary to carry out the task for which they are designed. Examples of this abound, each with its own personality and morality, from the Finnish mill of plenty, named the Sampo, and the Arthurian Excalibur, to the malevolent intelligence of Tolkien’s One Ring and Moorcock’s Stormbringer. Just as these enchanted mills, swords, and rings are capable of more than just their functional specification, so our connected devices, or Things, have processing and communicating capabilities well beyond the needs of the average lamp, umbrella, or bubble machine.
We began by looking at some examples of the Internet of Things in action. Throughout the book, we discuss many similar projects, from the perspective of creating the first prototypes in Part I to the extra effort required to manufacture and distribute them commercially in Part II. The Internet of Things can be characterised as joining the physical object, the computer embedded into it, and communication and code on the Internet itself. We focus on these three elements in the upcoming articles. We compared Internet-connected devices to enchanted objects.
I hope you found useful from this article and please share your feedback in the comment section.