Even before operators around the world are making anything much like significant revenues from the Internet of Things (IoT) market, the scope of the sector has already exploded. In the future, everything -- we are told -- will be networked.
But will that actually be the case? I think that, in reality now, and very probably long into the future, what we are looking at with the IoT sector is actually an Islands of Things market: A bunch of specialist devices that connect to one network and certainly don't talk to each other.
But first, some history!
"I could be wrong, but I'm fairly sure the phrase 'Internet of Things' started life as the title of a presentation I made at Proctor & Gamble (P&G) in 1999," British entrepreneur Kevin Ashton wrote in RFID Journal in 2009.
You'll note how sensor expert Ashton describes his original concept in the piece: "If we had computers that knew everything there was to know about things -- using data they gathered without any help from us -- we would be able to track and count everything, and greatly reduce waste, loss and cost."
Fast forward to 2016, the concept of IoT is far more focused on the sheer number of devices that will be connected by 2020 than talking about what to connect to, and what data to collect. Ericsson AB (Nasdaq: ERIC), for instance, is currently predicting 26 billion connected devices online by 2020. (Note: Even this is a revised estimate, down from 50 billion.) Never mind the quality, feel the width right?
There’s a simple reason for this, mobile operators the world over are hoping that revenues from IoT will replace declining voice spending and bolster data income in a hugely competitive market. So 10 billion, 20 billion, 50 billion connected devices makes a comforting number for operators that vendor sales and marketing departments can spin out in glossy marketing material.
In reality, the Islands of Things sector is already so fragmented that it is hard to really credit many of the exuberant predictions being thrown around. For starters, even being "connected" could mean a multitude of things: from a sensor that uses WiFi or ZigBee to transmit send data back once a week, to a connected car constantly sending data over a 4G LTE network.
In fact, the preferred connectivity standards for cellular IoT are not even set in stone yet. Most carriers today use low-power, low data-rate 2G (!) connections. Equivalent "lite" versions of the LTE standard have just been completed and will likely to be tested late this year for deployment in 2018. Meanwhile, there are a plethora of third-party Low-Power Wide-Area Network (LPWAN) specifications that IoT-curious carriers can deploy in the meantime.
Here's a handy list of (most of) the different networking protocols around IoT. Then consider that there are also multiple operating systems swirling around IoT!
It should be clear by now that IoT is, in fact the gumbo recipe of the wireless world: People might be talking about a few common elements when they say "IoT, " but one cook's dishes could contain a lot of different ingredients than another chef might select.
I don't think that's going to matter too much initially in some of the sectors where IoT projects are taking off now, such as agricultural monitoring, or sensor networks for utilities. Carriers and businesses can clearly build such proof-of-concept projects using the specific hardware, operating system and network that suits their purposes. It's when an agri-business needs to interface its sensor networks with a third-party drone crop-dusting company down the line that things get more complex.
Some sectors, meanwhile, clearly need solid cross-company and even cross-border standards to thrive. If the dream of streams of self-driving cars is to be realized outside of a few autonomous vehicles on the road then solid international standards need to be in place: Cars will need to connect to whatever network is available, be aware of other vehicles on the network, send diagnostic reports and hand-off between base stations at high-speeds. (I've sat through presentations about autonomous cars that has them driving at James Bond speeds down the Autobahn. Catch me in one of those? Not going to happen unless the network and hand-off is rock-solid!)
The Smart Home is the other place where compatibility -- and probably homogeneity -- appears very necessary. Apple and Google are already vying to be the king of the new digital home, in this case it just be that users will end up being an iOS or Android home: From smartphone, to tablet, to Internet-enabled kettle.
On the other hand, with reports of state actors and others carrying out ever more audacious hacks appearing weekly, we need to ask how connected to the Internet, some Industrial IoT networks actually need to be. Energy and water supply networks being hacked with nefarious intent should be a very real concern as more systems get ever-more connected.
So, rather than blandly talking of billions of networked devices making up the Internet of Things, I think we need to take a look at all these different Islands of Things that are now out there, and ask which need to be connected, to which networks and for what reasons. It might help us all to get a handle on the real scale of IoT and what actually needs to be done to make it happen.