The Internet of Things (IoT) is the foremost tech topic of the day. And for good reason: By connecting the unconnected through sensors and networks, and consolidating and analyzing the data generated, the IoT offers truly world-changing benefits for all verticals in all segments. Already, IoT technologies are helping to reduce spoilage in food, increase efficiency in factories, office buildings and homes, decrease downtime for industrial machines, improve our health. The list is nearly infinite.
However, harnessing the power of the IoT takes planning and preparation. You can’t simply go “buy some IoT,” flip a switch and be ready to go. Organizations considering deploying an IoT strategy should prepare using four key considerations:
1) Cut through complexity
As you begin to take a closer look at the IoT and determine how your organization can leverage connected devices and sensor data, the first thing you will face is daunting complexity. There are myriad data formats, few clear standards, and severe interoperability issues between operational technology (for example, machines on the factory floor) and IT. New security concerns have emerged as well, about digital damage and physical damage (Consider the havoc a factory robot could cause if hacked.) But all of this is moot if you can’t define your use case and how you plan to measure success. Failing to do so can cause your IoT project to stall as questions and concerns start to overwhelm answers and the organization loses interest and cuts funding. To complicate things further, there are currently over 400 IoT vendors, all promising you vast success, and finding the one -- or ones -- right for your organization is challenging. No one on vendor can do it all; look for the vendor that promises to work with you and partners to get it all done.
2) Make your data useful
At first glance, IoT can seem relatively straightforward: You stick sensors on things, connect those sensors into some IT application and then sit back and let the data flow into your analytics engine, fueling a stream of valuable insights. Oh, that it were so simple. But the fact is that data in its raw form is to insights as petroleum in its raw form is to premium unleaded. One is useful, the other not so much. So, how do you, in effect, refine your data? It’s a process: first you aggregate it by pulling from the relevant sources. Next you normalize the data so IT systems can understand it. Then, with your data aggregated and normalized, you can integrate it and finally analyze it for actionable insights. And depending on your analytics needs, you might have to do all of this very, very fast.
One more thing about data: In IoT environments there are many situations in which not every last bit of data is relevant. Consider a light bulb saying, “I’m on, I’m on, I’m on, I’m on.” Not important. But what about when the light bulb says, “Weird, I just dimmed for a split second.” Well, that could matter a great deal and you want to be aware of that. So, as you think about how to handle IoT data, you want to be clear on what data is important and be efficient in your approach to storage.
3) Architect for your analytics
IoT is not ‘one size fits all,” especially with regard to analytics. As you define your IoT strategy you want to be sure that right up front you know what you’re after in terms of analytics and then architect for your defined goal. Are you interested in edge analytics with an emphasis on streaming data for machine learning, for example on an assembly line? Or are you more interested in using data to make longer term strategic decisions? Or some combination of the two? You need to know before you start, because your goal will inform your architecture. It’s that simple.
When defining use cases, understanding the goal of analytics is important. There are four main types of data analytics. To illustrate these, I‘ll go back to the example of the light bulb: first up is descriptive analytics, which tells you that your light bulb is dimming; second is diagnostics, which reveals why your light bulb is dimming; third is predictive, which tells you your light bulb is about to fail; and fourth is prescriptive analytics, which tells you what you need to do to prevent your light bulb from failing. As you move along this progression of analytics, the sophistication of the underlying infrastructure and software increases dramatically, with prescriptive being by far the most demanding and difficult to achieve, but also offering potentially the greatest benefit.
4) Secure opportunities
Finally, security. At every step of your IoT project, security should be top of mind because every new connection brought forth by the IoT is a potential vulnerability. Compounding the problem is that—unlike IT, which has battled cyber threats for decades and has not only developed sophisticated defenses but also a mindset of healthy paranoia—the world of operational technology (OT) has with networks that historically have never touched the internet, processors running specialized operating systems and data protocols that are completely foreign to IT systems. While a measure of paranoia is good, being paralyzed is not. You must move forward: The advantages far outweigh the risks. And with the right security measures, you can more accurately measure risk and make the decision to act rather than let fear hold you back and create the biggest risk of all: standing still while your competition moves forward.
The hype surrounding the IoT is real, and so is the opportunity.
Before I wrap this up, I just want to acknowledge that the hype surrounding the IoT is at a peak so at the risk of being redundant, here are just a few statistics that reveal the undeniable opportunity the IoT promises: 30 billion connected things by 2020 (IDC); an aggregated value and economic benefit in excess of $1.9 trillion in the year 2020 (Gartner); a mind boggling increase in data from which insights can be harvested (EVERYONE). Bottom line, this is going to be big and every indication is that the risk of waiting on the sidelines far outweighs the risk of moving forward and establishing an IoT strategy for your organization. Moreover, research supports that if you approach the IoT by planning, and preparing your efforts can be richly rewarded.
Jeff Shattuck is the global messaging lead for the Internet of Things at Dell EMC. You can follow Jeff @jeffshattuck.