Putting data at the centre of IoT
Jim Morrish
Everybody knows that the IoT is all about the data. But I’m not so sure that that concept has really been fully thought through to its logical conclusions. In this blog, I’ll try to put ‘the data’ right at the core of the IoT.
So what’s new? Well, current thinking would run along the lines that within the IoT ‘data’ will be made available by an ever increasing number and diversifying range of IoT solutions. The secret sauce of the IoT kicks in when different types of data are combined in new and creative ways to realise some kind of valuable output.
That could be an increase in cost efficiency, or a whole new business model. Or potentially a new sector of the economy. What underpins this world view is a model of the IoT that is essentially still silo-based: a simple evolution from the old days of M2M, says Jim Morrish.
Sure, there’s a recognition that data needs to be released from these silos, and made available within analytics and other business and IoT platforms that draw inputs from as wide a range of sources as possible, but the conceptual architectural model is very much influenced by where the ‘things’ are, how they connect, and how data is made available to new systems and platforms, etc.
The fact is that today’s generally widely held concept of the IoT is rooted in an appreciation of today’s IoT architectures, which themselves are heavily influenced by legacy IT architectures.
So let’s unhinge ourselves from the current day, and consider how data might really be put at the core of the IoT. It strikes me that we would then arrive at a model of the IoT that is more characterised by a data stream. Data exists, it ‘is’, it flows through systems and applications, and that’s the real IoT. It’s independent of technical architectures and even applications, and is characterised only by SLAs and SLOs associated with different data.
Wrapped around that data stream would be a range of applications which connect into the stream, extracting the information that they need and then publishing their outputs back into the stream, or initiating other actions as appropriate.
In this worldview, a database simply becomes an application that stores data, making it available again at some later time on request. And edge computing is about getting access to key pieces of data either faster, or more reliably, or at a lower cost.
It’s a world where IoT applications move beyond the constraints of depending on the exhaust data from other IoT applications to augment their ‘native’ functionalities and capabilities, and are able easily to tap into primary data sources to construct new bespoke refined data feeds.
What would be the implications of this perspective on data on the established technical architectures of today’s IoT? Well, it remains to be seen, of course. But one consequence could effectively be the collapsing of the whole IoT platform stack down to the network level.
IoT platforms could effectively exist in fog/ edge environments. If the story of the IoT really is about the data, then it has to be about a data flow, and that data flow has to live in the network and all applications have to somehow tap into that data (and make their own outputs available).
All this leads inexorably to a conclusion that the IoT will be characterised by a truly intelligent network. Digital transformation through the use of distributed intelligence, rather than simply an internet of things.
The author of this blog is Jim Morrish
About the author:
Jim is a respected IoT industry expert, with over 20 years’ experience of strategy development and operations management. He is chairman of the Industrial Internet Consortium’s Business Strategy Task Group and led the development of the IIC’s Business Strategy and Innovation Framework. He is also co-chair of the IIC’s Business Strategy and Solution Lifecycle Working Group, and is responsible for the addition of the term Low Power Wide Area (LPWA) networking to the lexicon of the IoT.
Comment on this article below or via Twitter @IoTGN