Earlier this year, Angel Trains, the train leasing firm, demonstrated edge computing on its smart trains for the very first time. The project, which is being funded by a grant from the Department for Transport, hopes to improve the quality of experience of its customers on board trains by minimising the demand on the external backhaul network.
By deploying just a few devices per train carriage, Angel Trains will have created a self-managing network which, amongst other things, will provide an improved wireless network for its customers - with only an intermittent backhaul connection to the internet - by bringing edge computing as close to the user endpoint as possible. This, it seems, is only the latest example of the burgeoning new move towards edge computing globally.
For the past decade, cloud computing has been one of the most influential technological trends, but, as more businesses plan for large-scale IoT deployments, edge (or "fog") computing threatens to improve and replace it in 2018.
Ironically, edge computing has been bolstered by the popularity of cloud computing in the first place and, although the two can be said to have a symbiotic relationship, edge computing will shift much of the critical processing away from cloud data centres and, instead, over to the devices connected to them.
Edge computing is so called because it allows for real-time analysis of the data collected at the edge of a network. The term edge computing refers to the "edge devices" that lie at the outer rim of a network; for instance, your smart meter lies at the edge of your Wi-Fi network and your mobile phone at the edge of your carrier's network.
Using edge computing allows organisations to offer faster response times, reduced costs, as well as more bandwidth, and is increasingly important for any organisation wishing to stay ahead of their competitors with a comprehensive IoT strategy - especially in the channel.
With all the new devices connecting to any given network - whether it's due to bring your own device (BYOD) or IoT - there is now a heavy need for more intelligent devices and applications that can process the data collected and make critical business decisions directly at the edge of the network.
In order to gain any insight from the data collected it needs to be analysed, and the fastest and most logical place for this to happen is in the device itself. This then reduces bandwidth significantly, as the computing process happens in the devices itself and not, as is currently typical, in a central data centre where the information must then be sent back out to the edge device.
All of this newly acquired data needs to be stored and, increasingly, traditional storage methods are being complemented or replaced by software-defined storage (SDS) - a software layer independent of the hardware it overlays that provides the necessary provisioning for the management of data storage. It is often linked with hyper-converged infrastructure as this includes SDS and is predicted by Gartner to grow at 26.6% CAG over five years.
The digital economy is driving growth in many new, technological areas and thereby pushing organisations to change their business models to meet customer requirements. Trends like the gig economy (e.g. Uber, Airbnb) mean organisations' goods and services can now be offered on-demand, so the computing power required to run a business needs to be scalable and flexible. To remain competitive, organisations are increasingly able to deliver computing power as needed in a cloud-based model, only paying for the services they consume.
All of this represents a dramatic shift for the channel. It has never been more important to find partners that are transforming their business to capitalise on these new opportunities by migrating their organisation away from traditional, core technology requirements and providing future support for next-generation technologies and service requirements like IoT, security, cloud and analytics.
Steve Cant is director of enterprise infrastructure & comms Europe at Tech Data
