Missing 41bn IoT devices: Is this the biggest prediction miss in IT history?
Analysts and major enterprises predicted there would be 50 billion Internet of Things (IoT) devices connected by 2020, yet the true figure stands closer to 9 billion. Here, Eseye asks if this is the biggest prediction failure in the history of IT, and identifies the six key challenges that IoT must overcome in 2020 to reach its potential.
Compared to industry forecasts, 41 billion IoT connected devices have failed to materialise, says Eseye, a specialist in ubiquitous global IoT connectivity. In 2010, Ericsson predicted that 50 billion devices would be connected by 2020, a prediction echoed by Cisco in 2011. Yet, despite the enthusiasm for IoT,current estimates put the true figure for connected devices closer to 9 billion, with many of those being mobile phones.
Eseye has subsequently identified six key challenges that IoT must overcome in 2020 to reach its potential. (Click here to see Nick Earle’s Quickfire Video interview on this with IoT Now).
It’s now clear that successful IoT deployments are much harder than previously thought and substantial complexities have been glossed over. This is borne out by recent research from Cisco Systems which has found that more than 75% of IoT deployments fail. A lot of the damage happens before the devices even go live, however. Microsoft estimates that 30% of IoT projects fail at the Proof of Concept (PoC) stage, while Gartner says eight out of 10 IoT projects fail before they are even launched.
From hardware design and testing to connectivity, data management and global technical support, there are many obstacles to overcome.
The six challenges for IoT to overcome in 2020 as identified by Eseye, are:
1. Hardware needs to become relevant again
In IoT deployments 80% of the data and processing is at the ‘edge’ of the network. This is where the ‘things’ and sensors are and where data is captured. However, to make sense of it all, without the expense of having to back-haul the data into the heart of the network, it needs to be processed on the edge. To deliver successful deployments organisations need a strong understanding of how to optimise IoT hardware from circuit boards to firmware.
2. Bundled silicon to speed up deployment
The incorporation of secure IoT connectivity into silicon at the point of manufacture will go a long way to streamlining the IoT deployment experience. Bundling IoT capability at the silicon level significantly simplifies the setup and deployment of IoT devices. The real game-changer is that once the device is activated it should automatically connect to any network in the world, providing as close to 100% coverage anywhere in the world, and start provisioning data to either their on-premise solution or any one of the hyperscale cloud providers.
Eseye and Gemalto recently launched Intelligent Cloud Connect, in response to this challenge. The solution enables customers to develop and manufacture a single IoT product stock keeping unit (SKU) for any application, which then connects out-of-the-box to any mobile network in the world, while offering seamless and secure data provisioning to AWS IoT Core.
3. Localisation of devices to maintain uptime
A key precursor to the widespread adoption of IoT is the ability to quickly and simply connect devices anywhere in the world. Some suggest this exists through global roaming, yet a growing number of MNOs (Mobile Network Operators) and regulators are implementing permanent roaming restrictions which could mean that after three months an IoT device could be taken off the network.
For IoT customers with fixed devices around the world, this may result in the inability to use some networks beyond the short term unless they use a localised eSIM. As roaming falls out of favour organisations will need to turn to global ‘super’ Mobile Virtual Network Operators (MVNO’s) whose strategy is to localise connectivity wherever possible, in order to effectively deploy IoT devices anywhere in the world.
4. Overcoming the consumer centricity of
The Embedded Universal Integrated Circuit Card (eUICC) – often referred to as eSIM – was going to negate the need to migrate profiles between SIMs, enable everybody to work together and open a world of opportunities. Unfortunately, it has created a number of implementation challenges for business IoT use cases, as it was designed with the end consumer in mind, rather than industrial users.
Profile management and network switching must be driven by service provision rather than the profitability of an MNO. The best way to do this is for the profile management algorithm to be implemented in an abstracted and MNO agnostic switching platform – not by the MNO’s platform. Only by doing this can a single pane of glass management capability, single global invoice and single support service be delivered for an Enterprise’s total global estate of IoT devices.
5. Utilising Hypercloud
Cloud had its challenges with security. One of the biggest risks in IoT is the edge of the network and the massively expanded threat perimeter. The good news, however, is that several hyperscale cloud providers, such as AWS, are deploying standard security managed services features that audit the configuration of devices, monitor connected devices and detect abnormal behaviour to mitigate these security risks. Eseye predicts that 2020 will see at least 40% of new large IoT projects deployed in a hyperscale cloud platform.
6. The need for a ‘Star Alliance’ Federation model for IoT
With MNOs under pressure from complexity busting hardware, and with increased pressure on their commercial models, there is a growing need for them to compete for and deliver global IoT projects. A more commercially favourable and collaborative approach is required, such as a ‘federation of MNOs’, or to put it another way the ‘Star Alliance’ of the airline for IoT to prevent them from becoming a commodity.
Much like the Star Alliance where you would buy a single air ticket, travel around the world and pay once, if one MNO sells a global deal in one country, each MNO in the federation needs to then deliver the traffic requirements in their own country.
Nick Earle, CEO of Eseye, comments: “41 billion missing IoT devices is a monumental miscalculation and arguably is one of the biggest misses in the history of IT forecasting! The two questions that must be asked now, however, are why this happened and what can be done to rectify it? In our view, the ‘why’ primarily comes down to organisations underestimating the complexities of IoT deployments.
“The intricacies involved in creating specialist IoT device hardware, establishing access to global connectivity and the ability to manage vast amounts of data effectively and efficiently, are just some of those complexities. This was recognised by Gartner when it predicted in 2018 that 75% of IoT projects would take twice as long to deliver. Removing the complexities and barriers to IoT development and deployment will see a greater uptake on a global scale. We predict that 2020 will be the year that global IoT rollouts will take off, provided our six key challenges to IoT success in 2020 are overcome.”
To read more about the future of IoT in 2020 download the Eseye whitepaper here.