This article is courtesy of TechRepublic Premium. For more content like this, as well as a full library of ebooks and whitepapers, sign up for Premium today. Read more about it here.
The IoT isn't just about your refrigerator. Andrew Brust offers a level set on the industrial potential of the Internet of Things.
A lot of people see the term Internet of Things (IoT) and respond with a combination of confusion and skepticism. Even for technologists, the term conjures images of refrigerators and toasters dialing up a mothership computer at their manufacturer's headquarters and beaming telemetry data to it about what people are eating and when. That's at least a bit invasive. Maybe it's also harmless. But to many, it hardly appears important or useful; rather, it seems fed by a compulsion to amass increasing volumes of data and a default belief in the virtue of that quest.
It's industrial, stupid
While much of that view is understandable—and may in fact be quite correct—that consumer scenario is not really what IoT is about. Rather, IoT is more interestingly applied in industry, where it can be used to collect information that's always been needed but has until recently been estimated, based on hunch, instinct, and/or conjecture. Now, with advances in analytics, computing power, storage economics, and the ubiquity of internet connectivity, it's possible to substitute such assumptions and guesses with bona fide data.
The power to collect real data from devices, and to do so without an obligation to do so sparingly, has emancipated interesting analytics use cases across a number of industries and horizontal scenarios. It's also created a plethora of new standards, the challenge of unifying them, and a need to handle the storing, querying, analyzing, and modeling of this data in real time. Certainly, there's a lot to learn about, understand, and build. Seemingly, there's a lot of benefit to be had as well, though judging the validity of that belief requires some investigation and critical thinking.
I'll be writing about IoT for Tech Pro Research on a monthly basis, and there are a lot of IoT subdomains to dive into. But before drilling down into the particulars, a level set on some important IoT scenarios makes a better place to start. That's what this month's report focuses on.
Setting the context
Enjoying this article?
Download this article and thousands of whitepapers and ebooks from our Premium library. Enjoy expert IT analyst briefings and access to the top IT professionals, all in an ad-free experience.Join Premium Today
IoT is relevant in a number of use cases. It's extremely useful in manufacturing, for example. Getting telemetry data from equipment used to manufacture items is important, whether those items be consumer packaged goods or automobile parts or pieces of furniture. Manufacturing is a complex process, and complex processes, when run at scale, create huge potential for inefficiency.
Further, a process, by its very nature, can be discrete and closed, so getting insight into how things work within it can be tricky. IoT helps crack open the process black box and provides an understanding of what's going on. In so doing, it replaces assumptions with observations and deconstructs complexity into a collection of simple phenomena that, when measured, can be well understood and managed. IoT can remove mystique and opacity.
Services provided by state (or provincial) and local governments can benefit from IoT as well. How does the provision of public services get decided? How can such services be allocated most equitably and most effectively, so that they benefit the public and increase quality of life? How can such initiatives make a community more attractive, such that it appeals to new residents, increases its tax base, and completes a virtuous cycle of being sufficiently well-funded to provide even better services?
What about fleet management? How can a trucking or a taxi company best understand the optimal scheme for dispatching and loading vehicles in particular locales and domains? How can it minimize fuel costs, trip times, and on-time guarantees? How can insurance companies understand what habits lead to safe driving behavior and which customers exhibit those behaviors? IoT (in the form of vehicle telemetry) helps here too. Just think of how UPS, Uber, and companies like Progressive Insurance use data collection to help themselves through these operational management tasks.
IoT also runs hand-in-hand with predictive analytics. If you can understand when and under what circumstances inefficiencies or failures occur, that puts you in an excellent position to forecast and prevent such events as operations are observed. Manufacturing organizations use this to predict when a piece of equipment may be nearing a breakdown and can remove that equipment from service preventively, thus avoiding a breakdown in production and the disruption and costs that would result.
Customers and vendors at odds?
There are, of course, applications on the consumer side too. Connected devices and appliances in so-called "smart homes" do provide useful data. Wearable technology does as well. However, as you consider these scenarios, it's important to understand to whom the data is useful.
If a refrigerator can detect when certain items or supplies are running low and automatically order more, that can help the consumer. If a thermostat can understand under what circumstances (times of day, weather patterns) use of energy tends to be high—and can advise on what times it's more or less expensive to be home or be out—that can benefit the consumer too.
But if these same devices are providing telemetry that helps the vendors behind these devices sell more of them, or send a targeted list of consumers to grocery stores, distributors, or brands, it's second and third parties who benefit. That's not necessarily evil... but it's different from providing direct benefit to the user. And such use of telemetry data may create some ambivalence around data sharing and the importance of IoT in general. So policy and ethics matter; they have business impact.
If a machine that administers dosages of chemotherapy can report on its own operations and that data can be correlated with data from medical equipment monitoring the vital readings of the patient to whom the meds are being administered, that clearly benefits the patient. If that same chemotherapy administering equipment is repurposed to dispense soft drinks, the telemetry data benefits the beverage manufacturer more than the person drinking the beverages.
You already know IoT, even if you didn't know that
A key to appreciating the value of IoT is recognizing that it's not really new; what's changed is the variety of places where the technology can be applied. But if you consider that more limited applications have existed for quite some time and offer utility and value that is already well understood, accepting the utility of IoT can involve a much smaller leap of faith.
For example, think of all the data that has been gathered for years at cashier's stations in retail establishments—that is, at points of sale (POS). Understanding buying habits, which purchases tend to accompany others, peak shopping times, and even personnel productivity can result from careful collection and analysis of POS data. Now consider that all IoT really does is extend such data collection to other locations, beyond points of sale. IoT begets data collection at "Points of Everything," if you will.
Other well-known examples include software, mobile apps, and Web sites, many of which record event-based information in log files. This is a standard practice, born of the need to perform diagnostic analysis in the case of bugs in the software or operational anomalies. What if such logging could become standard practice everywhere... and not just for software but for hardware too?
The process of inserting code that performs logging is called "instrumenting." IoT, then, can be thought of as the instrumentation of processes, equipment, and even people. Such instrumentation can provide the same analytics-based understanding of how things work that software defect diagnostics, Web analytics, and mobile app analysis provide in their own domains.
IoT is really just a three-letter acronym for a way of approaching the management of physical objects in industrial and consumer scenarios. Just because something is physical doesn't mean it has to be analog anymore. Tangible objects can be observed, measured, and analyzed though digital means—so physical interactions can become transactions, in the database sense.
What's on cable?
A great example here is the humble cable TV converter ("cable box"). I remember when my family first got cable TV in the early 1980s; our converter box actually had a tuning knob on it, just like older televisions did. It could tune into channels 2-13 and A-W (yes, cable channels in Manhattan were at first lettered, rather than numbered). Tuning in a cable network like ESPN or CNN actually involved turning a dial. That's about as physical as it gets.
Today's cable boxes are, essentially, PCs. They contain digital, addressable converters; they may have hard drives to support DVR functionality; they have on-board software with high-definition user interfaces that run on full-fledged operating systems, which, when started up, go through a true (and rather long) boot sequence. Every change of the channel, and observation of which networks are viewed when, can be recorded. If a particular show or advertisement coincides with a subscriber tuning away to something else, that can be observed too. So can the number of hours of programming viewed on particular days, viewing of on-demand versus linear programming, and other such details.
While the IoT does not necessarily bring about such extreme transformations of physical objects from analog to digital as cable set-top boxes have undergone, it brings in at least some element of that. Whether through a complete redesign of an object or more of a retrofit, it is evidence of an overall transformation of how "things" are managed.
And because cloud infrastructure is widely used in IT today, the industry has natural collecting points for all this data. It can that be overwhelming, but this use of cloud data centers is not much different in principle from having an engine room, in a vast ship or large office building, where assets are monitored and from which communication is managed. What's different is that the barriers to entry of having such a control center have been lowered.
Viewed in this light, and combined with the relative ease of outfitting objects of all kinds with sensors, the ubiquity of internet connectivity with which those sensors can report their data, the relatively low expense of the disk space needed to store that data, and the technological advances that make capturing and analyzing that data in real-time feasible, it's no wonder that IoT has become a hot area. And with our context now set, there's an awful lot to explore and understand more fully.
In future reports, we'll look at different scenarios where IoT is making a difference, in helping organizations do things they couldn't do before and allowing them to do things faster, less expensively, and more efficiently than they could do otherwise.