Moving aggressively from early adoption toward supporting the world
We find ourselves at a seminal moment in the evolution of information technology: The stuff around us, even attached physically to us, is getting its own brain, and voice. Thanks to the ever-steady progression of Moore’s Law, as applied not only to microprocessor power but also to memory, storage and bandwidth — specifically, wireless bandwidth — plus other key commercial factors, we’re seeing almost anything and everything now capable of telling its own story. Generally, this is being lumped together as the Internet of Things (IoT), and we’re clearly in love with it. Couple it with cloud computing, big data analytics and a few other already ubiquitous buzzwords, and you’re able to imagine the day when your healthcare provider conspires with your refrigerator and treadmill to adjust insurance premiums based upon your lifestyle. And don’t forget the coming of self-driving cars, all the rage.
But let’s talk about the forces that have historically driven major technology advancements: business and governmental applications. When viewed through the lens of business, the IoT has been here in various forms for a long while, perhaps not beaming information directly over IP stacks and using open wireless protocols, but definitely present. Logistics, supply chain management, transportation and many other facets of what runs the real world, are all “IoT” in some form. The same goes for closed systems such as battlefield operations and what happens inside of factories. All of this presents a conundrum for forward-thinking IT architects — how to morph what’s already out there, humming along in many, many silos and inside of electronic fences, towards support for standards, scalability, data sharing and other emerging requirements, whilst protecting their functionality and whatever proprietary advantage they were installed to provide in the first place.
First, we gather
Typically, when a new, sweeping technology shows up, we look around and see who else has noticed and started to care. This inevitably leads to meetups at conferences, published papers and the formation of industry- and/or government-driven standards groups. Most of these in the end don’t produce anything that gets adopted widely (witness, if you go back, the triumph of TCP/IP over OSI…), but they generate a lot of attention and activity while going about the “standardization process” they are intended to pursue. There is a lot of this happening now with IoT, and some may actually yield beneficial results. Most of them are focused on the areas of device-level communications protocols and interoperability, a few on file interchange formats, some more on marketing IoT. The latter tend to be promulgated by vendors that have built an “IoT platform” or, in many cases, repurposed something they built for another application, and now being repositioned to take advantage of the hot trend that IoT represents. All good; there is a grand history of such efforts within the computing and communications industries. Again, the majority don’t yield results measured in financial returns. So how do we get to the real meat of this new area, being the business advances that it promises to enable?
Next, we ponder
According to my friend McKinsey partner Mark Patel, a significant portion of information that is already being collected via IoT-type systems and sensors is either used for an instant and discarded, or stored in a corporate archive, never to be seen again. “In one recent industrial example, we found that of 30,000 different sensor and machine inputs, only 1% of the data was actually being analyzed,” says Mark. In other words, organizations put in sensors to gather data from whatever it is that their business is concerned with, look at it (or don’t), make decisions (sometimes), and that’s the end of it. Ponderous, indeed. So with the advent of routing such data over standardized secure protocols, allowing it to bubble up to analytical systems, possibly even getting it to third parties for further potential use, can we make IoT data truly useful to business? Mark thinks so: “Our recent research indicates that over $4 trillion of untapped value could be addressed by 2025,” he says. (Take a look at a huge study that he and fellow McKinsey colleagues have released. It tells a big story.) To have so much data sitting around unmined seems illogical, but if you look at the historical reasons why the data was generated in the first place, it’s clear that companies had only one purpose in mind.
My current favorite example of repurposing IoT data that was collected for one purpose and mined for a completely different and unrelated application occurred last year, when the Earth shook: during a relatively minor ground temblor in Napa Valley, California, there were enough people wearing fitness trackers that data scientists could tell you roughly where the earthquake occurred and its intensity, by mining the spike in heart rates and when scrutinized in detail coupled with geolocation data, it gets pretty close to the USGS findings on the same incident. This example really opened my mind to the possibilities of mining all sorts of IoT data, for uses never imagined by the progenitors of the devices generating it.
Then, we pursue
The business world, depending on the industry, has a history of cross-correlative analysis that has yielded tremendous benefits and often competitive advantage — the use cases are many and varied, far too many to enumerate here. But extrapolate that using the incredible amount of physical-world data generated now and going forward via IoT, and it gets very exciting indeed. The fitness tracker example, perhaps most notable because it didn’t require knowing someone’s personal identifying information (PII) to be useful, serves to expand one’s thinking about how to leverage an organization’s investment in sensor technology and systems, along with wireless, cloud computing and big data analytics. We’ll find ways to bring these together for business benefit and eventually, to enhance quality of life. It’s pretty cool.