The internet of things wave is here, and its promise of flashy new gadgets is quickly being overshadowed by its promising effects on both the enterprise and industry verticals. However, at the end of the day, Don DeLoach, former president and CEO of Infobright (it was acquired by Ignite Technologies in March 2017), knows IoT implementation is all about the data and how that data is being used.
In his recently published book, The Future of IoT, co-authored by Gartner analyst Emil Berthelsen and Pentaho director of enterprise solutions Wael Elrifai, DeLoach explores the current state of IoT, where we are going, and why a more holistic approach will help us get there. Get a sneak peek here.
The book is called The Future of IoT. What is the future of IoT?
Don DeLoach: The future of IoT is rooted in widespread acceptance and implementation of IoT in all types of enterprises. But as that happens, the focus will shift increasingly to the enterprise, where leveraging the underlying IoT data will create the most value and, as such, the data primacy issues around IoT data will be a key consideration in deployment architectures in the future. So in short, the future of IoT is leveraging of the utility value of the IoT data at the enterprise level.
What's the main point about the future that readers should take away?
DeLoach: That the silo systems that mark the emergence of IoT, like smart lighting or smart refrigerators or industrial equipment, will give way to a more holistic treatment of the participating systems. The reader should take away that the deployment architectures will need to accommodate this holistic approach.
How far are we from this holistic treatment, and how do we get there?
DeLoach: I think we are beginning to see movement in this direction, but the progression will likely be slow. There are a number of platforms capable of supporting this type of approach (Amazon IoT, Hitachi Lumada, Verizon and BlackBerry, to name a few), but despite there being underlying technologies that can more or less accommodate the direction, the progress will be equally tied to the emergence of enterprises demanding more control over the data and product providers recognizing that while they don't want to 'lose control' of the closed loop silos, providing their enterprise buyers the ability to better control the data will increasingly become a competitive advantage for them.
I also think we will begin to hear more about enterprises and organizations that are taking these positions. So in short, I think we are not there yet and the progression will take a while, but we will begin to see that happening slowly in the near future.
In the beginning, you reference an article by PTC's Jim Heppelmann and Harvard professor Dr. Michael Porter on the 'five stages of IoT.' It was written a couple of years ago; where do you think we are in terms of these stages, and are we where we should be in terms of IoT implementation?
DeLoach: We are somewhat in between stage three and stage four from a practical standpoint. Stage three is the smart connected products stage. This is marked by the association of IoT with "IoT-enabled products" and the lens is through the product supplier. The fourth phase is product systems, which have products talking to other products, like what we are seeing in smart homes and certain product sets designed to work with other products. The main consideration here is the time it is taking for the market to evolve. While there are clearly examples of movement to the product systems phase and we are beginning to see some 'system of systems' examples, the majority of money is still associated with the more basic IoT implementation of consumer products. As people and organizations gain more experience, they will also more carefully consider the data they would like to use from these products and, more specifically, how they would like to use that data, especially in the context of data from other IoT products. "Where we should be" is considering where the market is going and beginning to plan our technology roadmap, training and hiring plans, and contractual frameworks around gaining the maximum leverage of the data. As an aside, this can be done while still providing the product makers all the data they are recovering now, but it doesn't need to be at the expense of the product user (ostensibly the enterprise).
The Future of IoT
The Future of IoT focuses a great deal on IoT data and the importance of using it properly. What is holding organizations back from reaping the full benefits of IoT data, and how do you think they will get where they need to be in the future?
DeLoach: The progression of IoT and the focus on IoT-enabled products is an understandable progression. But as long as organizations do not have full access to and control of the data, they will be unable to gain the maximum leverage of that data. I would argue that if the organization using the IoT-enabled products is contractually bound to have the product provider both dictate the specific data and the flow of that data, then the limitations on the enterprise are significant. So the key is the architecture. We highlight a concept called the 'first receiver' in the book. There is nothing overly complex about this. In fact, the notion of leveraging the utility value of the data is the basis for relational databases that have been around for decades. The main premise is separating the creation of data from the consumption of data. This is commonly associated with an event-driven, publish-and-subscribe architecture. That same temperature you are getting from the turbine might be helpful to the turbine maker, the turbine user, the turbine users' supply chain partners and even regulatory agencies like OSHA. It's still the same piece of data. The key is to get the right data into the right hands in the right way. That's where the leverage is realized.
Taking that a step further, where do we stand today in terms of ownership of data? What should organizations look for in regard to data ownership?
DeLoach: It's really mixed and there are clearly a variety of options on this. I also believe options are changing. Also important in this equation is the clarification between ownership and stewardship. One party may technically own the data while others may be given control. Think of yourself setting use permissions on certain apps. You may control (to some degree) the data without necessarily owning it. I think the enterprises that acquire and use IoT-enabled products should own and control the data but, in certain cases, allow for some control to be passed to other constituents to the extent it makes sense. I would also argue that the relationship between the provider of the IoT-enabled products and the user of those products should be contractually established so the product provider can continue to get the data from the products in order to service those products, as well as continue to make them better.
Is IoT, as a concept, too big to be useful? Should we narrow what we mean when we say IoT?
DeLoach: Conceptually, the broader the signature the better (at least in many cases). Consider population health. Understanding medical informationis critical, but marrying that information to environmental information, demographic information and even things like weather or crime could conceivably contribute to creating a pattern that reveals insight. The ultimate goal is to take more informed actions that lead to better outcomes, regardless of the vertical orientation.
We should recognize IoT has huge implications for how we will progress as households, organizations and as a society. I think the view of IoT being listed to the specific product view is a mistake, especially with regard to how the data is treated and leveraged. Likewise, the way the world will look five or 10 years out will most certainly be beyond what many could possibly consider right now. I think when we say IoT we should start by viewing it from the vantage point of the user of multiple IoT systems, and how those systems interact and how the underlying data can be used to create maximum value.
Should we stop talking about IoT implementation singularly and instead talk about IoT logistics, IoT automotive, IoT automation and IoT healthcare, all of which will be assumed to include lots of connectivity?
DeLoach: We probably should in many respects. The truly 'broad view' of IoT will reach beyond this, even when an organization is reaching beyond its normal scope for information that might seem completely unrelated, like a fast-food franchise owner trying to understand micro weather patterns or environmental data or concert schedules; that seemingly unrelated data would be used to enrich the baseline data of the operation. The franchise owner wants to sell more at a higher price while spending less. Understanding how the traffic flows relative to certain events, or what the point-of-sale data suggests when the weather patterns follow a certain trend, or some combination thereof, all enriches the signature used to take predictive actions to achieve a better outcome. In this regard, domain expertise and context make all the difference. So I believe the 'narrowing' has more to do with the data model and associated consuming applications. But that also accommodates the enriched signatures and provides the basis for gaining better and better insight for that particular enterprise.