If you haven’t already heard of Internet of Things, you soon will.
There is considerable business and technological momentum behind Internet of Things, also known as IoT. Gartner has put IoT way up there in its hype cycle; practically every major technology company is in the process of developing an IoT product; a number of universities in the United States, Europe and Asia have launched big R&D programs in IoT; the European Union is funding the massive Internet of Things Initiative; and China has identified IoT as a technology of national priority.
The proponents of IoT imagine a world in which billions of objects of various sorts (cameras, pacemakers, RFID tags, sprinklers—you name it) are connected to the Internet, communicating and cooperating with one another.
Why now? After all, this idea has been around for well over a decade under different names—object Internet and machine-to-machine (M2M) being two of the better known—and has occasionally been the butt of jokes (“Did you hear the one about what the toaster said to the refrigerator?”). So is this old wine in a new bottle? Or is this renewed interest based on some major new technological breakthrough?
As it turns out, it’s neither. Much as social networks came of age as more and more people got online, networks of communicating objects are proliferating as the world becomes filled with more and more sensors and other intelligent objects, supporting a broad range of applications. However, there isn’t any identifiable single technology called IoT that supports the wildly different scenarios lumped together under this term.
The purpose of this article is twofold. One is to present a range of representative examples to show that devices of various sorts communicating and cooperating with one another and with IT systems is already very much a reality. The second is to emphasize that to reap the benefits, one must get past four common myths about IoT.
A look back
In the mid-1990s, the World Wide Web was a network of linked documents. Search engines enabled you to find the documents, and links enabled you to navigate across them. But many technologists envisioned the Internet as much more than the Web. Could the Internet also be populated by not just documents but objects that could discover one another, establish connections, communicate and cooperate?
RFID—radio frequency identification—was also becoming more common at this time, lending credibility to this idea. RFID tags are microchips that emit a simple signal to identify themselves and can be attached to any object (so, by proxy, they can identify the object they are attached to). The scanner that detects the signal does not have to be in physical contact with the RFID tag, nor do the two have to be in each other’s line of sight. Technically, the contents of an entire freight truck or even a warehouse can be inventoried in a moment using RFID.
The technology got a shot in the arm when a major retailer, looking to build a considerably more efficient supply chain, announced that it would be requiring its major suppliers to tag their products with RFID.
Anticipating large-scale adoption of IoT, languages for device communication were proposed by technologists and consortia for object Internet were formed. Further, many experts opined that if each object had its own unique Internet address, the 4 billion or so addresses available under the Internet standard IPv4 would not be adequate.
Although this problem could be solved by other means (such as Network Address Translation, one of the most commonly used techniques today), a new standard—IPv6—was developed to expand the available number of Internet addresses.1
Why this history lesson? To underscore the point that the idea of IoT has a long history of failed technologies and promises in its wake. None of the proposed device communication standards gained any traction.2 The large retailer revised its mandate several times, and nearly a decade later, it implemented a much less ambitious RFID-based supply chain. Although IPv6 is more than a dozen years old, the Internet continues to operate merrily with the ancient IPv4.
So, judged by past experience, IoT would seem more like bad old wine in a new leaky bottle.
However, if we take a sober look at devices interacting with one another and with IT systems, we will find a wide range of applications today spanning multiple technologies and industries.
RFIDs may not have revolutionized the world, but RFID applications are ubiquitous nonetheless. Your car key has an RFID tag so that your vehicle could not be hot-wired and driven away without your particular key; your employee badge is RFID-based, and so is your dog’s collar; hospitals use RFID tags to recognize and track the movement of human and medical assets ranging from medical equipment to blood samples, even patients. On the more esoteric side, sheep farms in the Australian Outback use RFID to identify sheep when they enter feeding stations and, based on the sheep’s medical history, dispense the right amount of food and medication.
A little higher up in the food chain from RFIDs are sensors that can measure one or more conditions in their environment. Smoke detectors and motion detectors are now routine in safety and surveillance applications; sensors collect your toll when you drive. From banking to border crossings, biometric sensors are increasingly finding their way into applications that require secure authentication of individuals.
In the utilities industry, smart meters (also called AMI, or advanced metering infrastructure) measure a customer’s power usage every few minutes and help power companies with demand prediction, which, in turn, can help with peak shaving and valley filling to optimize the use of the power infrastructure. Sensors monitor equipment such as oil pipelines in remote locations to identify cracks, leaks, malfunctions and theft.
Motes (also called smart dust) are small, self-contained chips that can be scattered across a wide area and integrate one or more sensors with a radio and battery. They establish communication among themselves to create an ad hoc network; collectively, they can monitor their environment and raise an alarm when they detect an abnormal condition, such as a sudden temperature rise across a wide area. Besides defense applications, motes are being evaluated for detecting and monitoring forest fires.
The above is a smattering of applications enabled by smart devices, ranging from such mundane uses as identification and tracking to process control and adaptive decision making.
However, to leverage the power of devices communicating with one another to achieve business value, one has to get past four common myths and misconceptions about Internet of Things.
Myth 1: IoT is a technology
Although the term IoT is useful when referring to a wide range of applications that involve smart devices, you may have noticed that there is precious little in common—technologically—across the applications mentioned above. Even among the RFID-based applications (which can be replaced by any number of identification technologies, such as magnetic strips and biometrics), there is very little in common besides the RFID tags themselves.
In other words, IoT is a concept, not a single technology you’d buy off the shelf.
Myth 2: IoT is the next wave of the Internet
The term “Internet of Things” conjures up images of billions of objects freely discovering, communicating and cooperating with one another over the Internet.3 This is not only a fairly silly idea, it’s a dangerous one as well, especially when the devices can perform a physical action, such as opening a door, turning on a conveyer belt, or opening and closing valves in a hydroelectric turbine in response to a command that comes across the Internet.
A significant number of devices may use TCP/IP—the communication protocol used by the Internet—to communicate within proprietary networks, but that will be as close to the Internet as they’ll get. In fact, many of the devices—the RFID in your car key, for example—may not even communicate to any network at all and may use communication protocols optimized for a very specific application.
In other words, its name notwithstanding, IoT has nothing materially to do with the Internet.
Myth 3: Regulations on data privacy is a critical enabler of IoT
Since sensors can track people and their behavior, critics often claim that privacy regulations are a critical enabler for large-scale adoption of IoT. In my opinion, this is a holdover from the early days of RFID, when that technology was often mentioned in the context of consumer goods and retail.
While applications that directly touch consumers and citizens will indeed get a fillip from the enactment of satisfactory data privacy regulations, most applications of IoT will have little or nothing to do with consumers and data privacy. An oil company using sensors to monitor its Alaskan pipeline or a kennel club using RFID tags to locate lost canines or a power generation company using sensors on its turbines to predict and avoid potential failures—all applications in operation today—do not fall within the purview of regulations or data privacy concerns.
Even when data privacy is an issue, companies with imagination have always found ways to overcome their consumers’ privacy concerns by inventing new business models. A well-known American insurance company has been able to sign up volunteers who agree to have their driving habits profiled by a special device installed in their automobiles; the insurer, in return, offers lower insurance rates to good drivers.A credit card company, when presented with a transaction, uses location data from the customer’s cell phone and from the transaction to identify potential fraud.
In other words, companies holding off on IoT because they’re waiting for data privacy regulations are liable to be waiting for a long time.
Myth 4: IoT needs device communication standards
Something else you’ll hear a lot about in IoT discussions is the lament over the need for various device communication standards. Yes, standards are necessary if random devices have to talk to other random devices over the Internet. But, as pointed out above, most IoT applications are likely to be highly specialized, communicating over proprietary networks.
To be sure, standards never hurt and are particularly useful for applications that cross organizational and system boundaries (such as supply chain management). But I believe that organizational imagination focused on specific applications in specific industries will be a much bigger enabler of IoT than any sweeping standard across heterogeneous devices that perform a wide range of functions.
Today, many companies are reaping significant benefits from device-to-device communication, ranging from asset tracking and supply chain optimization to unique, industry-specific applications such as customized pricing.
As the IoT hype picks up—as I expect it will in the next year—common myths about IoT are likely to be perpetuated. But IoT is nothing more than a catchy name for a loose category of applications that involve identifiers, sensors and actuators working with one another and IT systems.
But if you really insist that your toaster talk to my refrigerator over the Internet, we can definitely use an old, well-known standard: Esperanto.
1. To be sure, IPv6 not only expands the address space; it also provides a much richer protocol for communication across the Internet. (Back to story.)
2. HP’s eSpeak had some traction in the auto industry, but is no longer supported. (Back to story.)
3. Estimates range from 50 billion to 1 trillion objects connected to the Internet by 2015. While I do not doubt the number of objects, I seriously doubt that any significant percentage of them will be connected to and operated via the Internet. (Back to story.)
About the author
Kishore S. Swaminathan is Accenture's chief scientist and the global director of Accenture Technology Labs' systems integration research. He is responsible for defining the company's vision for the future of technology and setting its research and development agenda. Based in Beijing, Dr. Swaminathan has spent his Accenture career researching cutting-edge technologies. Winner of the 2000 Computerworld Smithsonian award for the best application of IT, Dr. Swaminathan has worked on more than a dozen research projects and has as many patents to his credit.